Google Git
Sign in
ceres-solver / ceres-solver / 939253c208718689c8f141a2aa97ba0827478a93 / . / docs / source / nnls_modeling.rst
blob: faa3c2674bddd2a991f95a9ed2f292ba555a419f [file] [log] [blame]
.. default-domain:: cpp
.. cpp:namespace:: ceres
.. _`chapter-nnls_modeling`:
=================================
Modeling Non-linear Least Squares
=================================
Introduction
============
Ceres solver consists of two distinct parts. A modeling API which
provides a rich set of tools to construct an optimization problem one
term at a time and a solver API that controls the minimization
algorithm. This chapter is devoted to the task of modeling
optimization problems using Ceres. :ref:`chapter-nnls_solving` discusses
the various ways in which an optimization problem can be solved using
Ceres.
Ceres solves robustified bounds constrained non-linear least squares
problems of the form:
.. math:: :label: ceresproblem_modeling
\min_{\mathbf{x}} &\quad \frac{1}{2}\sum_{i}
\rho_i\left(\left\|f_i\left(x_{i_1},
... ,x_{i_k}\right)\right\|^2\right) \\
\text{s.t.} &\quad l_j \le x_j \le u_j
In Ceres parlance, the expression
:math:`\rho_i\left(\left\|f_i\left(x_{i_1},...,x_{i_k}\right)\right\|^2\right)`
is known as a **residual block**, where :math:`f_i(\cdot)` is a
:class:`CostFunction` that depends on the **parameter blocks**
:math:`\left\{x_{i_1},... , x_{i_k}\right\}`.
In most optimization problems small groups of scalars occur
together. For example the three components of a translation vector and
the four components of the quaternion that define the pose of a
camera. We refer to such a group of scalars as a **parameter block**. Of
course a parameter block can be just a single scalar too.
:math:`\rho_i` is a :class:`LossFunction`. A :class:`LossFunction` is
a scalar valued function that is used to reduce the influence of
outliers on the solution of non-linear least squares problems.
:math:`l_j` and :math:`u_j` are lower and upper bounds on the
parameter block :math:`x_j`.
As a special case, when :math:`\rho_i(x) = x`, i.e., the identity
function, and :math:`l_j = -\infty` and :math:`u_j = \infty` we get
the more familiar unconstrained `non-linear least squares problem
<http://en.wikipedia.org/wiki/Non-linear_least_squares>`_.
.. math:: :label: ceresproblemunconstrained
\frac{1}{2}\sum_{i} \left\|f_i\left(x_{i_1}, ... ,x_{i_k}\right)\right\|^2.
:class:`CostFunction`
=====================
For each term in the objective function, a :class:`CostFunction` is
responsible for computing a vector of residuals and Jacobian
matrices. Concretely, consider a function
:math:`f\left(x_{1},...,x_{k}\right)` that depends on parameter blocks
:math:`\left[x_{1}, ... , x_{k}\right]`.
Then, given :math:`\left[x_{1}, ... , x_{k}\right]`,
:class:`CostFunction` is responsible for computing the vector
:math:`f\left(x_{1},...,x_{k}\right)` and the Jacobian matrices
.. math:: J_i = \frac{\partial}{\partial x_i} f(x_1, ..., x_k) \quad \forall i \in \{1, \ldots, k\}
.. class:: CostFunction
.. code-block:: c++
class CostFunction {
public:
virtual bool Evaluate(double const* const* parameters,
double* residuals,
double** jacobians) = 0;
const vector<int32>& parameter_block_sizes();
int num_residuals() const;
protected:
vector<int32>* mutable_parameter_block_sizes();
void set_num_residuals(int num_residuals);
};
The signature of the :class:`CostFunction` (number and sizes of input
parameter blocks and number of outputs) is stored in
:member:`CostFunction::parameter_block_sizes_` and
:member:`CostFunction::num_residuals_` respectively. User code
inheriting from this class is expected to set these two members with
the corresponding accessors. This information will be verified by the
:class:`Problem` when added with :func:`Problem::AddResidualBlock`.
.. function:: bool CostFunction::Evaluate(double const* const* parameters, double* residuals, double** jacobians)
Compute the residual vector and the Jacobian matrices.
``parameters`` is an array of arrays of size
``CostFunction::parameter_block_sizes_.size()`` and
``parameters[i]`` is an array of size ``parameter_block_sizes_[i]``
that contains the :math:`i^{\text{th}}` parameter block that the
``CostFunction`` depends on.
``parameters`` is never ``NULL``.
``residuals`` is an array of size ``num_residuals_``.
``residuals`` is never ``NULL``.
``jacobians`` is an array of arrays of size
``CostFunction::parameter_block_sizes_.size()``.
If ``jacobians`` is ``NULL``, the user is only expected to compute
the residuals.
``jacobians[i]`` is a row-major array of size ``num_residuals x
parameter_block_sizes_[i]``.
If ``jacobians[i]`` is **not** ``NULL``, the user is required to
compute the Jacobian of the residual vector with respect to
``parameters[i]`` and store it in this array, i.e.
``jacobians[i][r * parameter_block_sizes_[i] + c]`` =
:math:`\frac{\displaystyle \partial \text{residual}[r]}{\displaystyle \partial \text{parameters}[i][c]}`
If ``jacobians[i]`` is ``NULL``, then this computation can be
skipped. This is the case when the corresponding parameter block is
marked constant.
The return value indicates whether the computation of the residuals
and/or jacobians was successful or not. This can be used to
communicate numerical failures in Jacobian computations for
instance.
:class:`SizedCostFunction`
==========================
.. class:: SizedCostFunction
If the size of the parameter blocks and the size of the residual
vector is known at compile time (this is the common case),
:class:`SizeCostFunction` can be used where these values can be
specified as template parameters and the user only needs to
implement :func:`CostFunction::Evaluate`.
.. code-block:: c++
template<int kNumResiduals,
int N0 = 0, int N1 = 0, int N2 = 0, int N3 = 0, int N4 = 0,
int N5 = 0, int N6 = 0, int N7 = 0, int N8 = 0, int N9 = 0>
class SizedCostFunction : public CostFunction {
public:
virtual bool Evaluate(double const* const* parameters,
double* residuals,
double** jacobians) const = 0;
};
:class:`AutoDiffCostFunction`
=============================
.. class:: AutoDiffCostFunction
Defining a :class:`CostFunction` or a :class:`SizedCostFunction`
can be a tedious and error prone especially when computing
derivatives. To this end Ceres provides `automatic differentiation
<http://en.wikipedia.org/wiki/Automatic_differentiation>`_.
.. code-block:: c++
template <typename CostFunctor,
int kNumResiduals, // Number of residuals, or ceres::DYNAMIC.
int N0, // Number of parameters in block 0.
int N1 = 0, // Number of parameters in block 1.
int N2 = 0, // Number of parameters in block 2.
int N3 = 0, // Number of parameters in block 3.
int N4 = 0, // Number of parameters in block 4.
int N5 = 0, // Number of parameters in block 5.
int N6 = 0, // Number of parameters in block 6.
int N7 = 0, // Number of parameters in block 7.
int N8 = 0, // Number of parameters in block 8.
int N9 = 0> // Number of parameters in block 9.
class AutoDiffCostFunction : public
SizedCostFunction<kNumResiduals, N0, N1, N2, N3, N4, N5, N6, N7, N8, N9> {
public:
explicit AutoDiffCostFunction(CostFunctor* functor);
// Ignore the template parameter kNumResiduals and use
// num_residuals instead.
AutoDiffCostFunction(CostFunctor* functor, int num_residuals);
};
To get an auto differentiated cost function, you must define a
class with a templated ``operator()`` (a functor) that computes the
cost function in terms of the template parameter ``T``. The
autodiff framework substitutes appropriate ``Jet`` objects for
``T`` in order to compute the derivative when necessary, but this
is hidden, and you should write the function as if ``T`` were a
scalar type (e.g. a double-precision floating point number).
The function must write the computed value in the last argument
(the only non-``const`` one) and return true to indicate success.
For example, consider a scalar error :math:`e = k - x^\top y`,
where both :math:`x` and :math:`y` are two-dimensional vector
parameters and :math:`k` is a constant. The form of this error,
which is the difference between a constant and an expression, is a
common pattern in least squares problems. For example, the value
:math:`x^\top y` might be the model expectation for a series of
measurements, where there is an instance of the cost function for
each measurement :math:`k`.
The actual cost added to the total problem is :math:`e^2`, or
:math:`(k - x^\top y)^2`; however, the squaring is implicitly done
by the optimization framework.
To write an auto-differentiable cost function for the above model,
first define the object
.. code-block:: c++
class MyScalarCostFunctor {
MyScalarCostFunctor(double k): k_(k) {}
template <typename T>
bool operator()(const T* const x , const T* const y, T* e) const {
e[0] = k_ - x[0] * y[0] - x[1] * y[1];
return true;
}
private:
double k_;
};
Note that in the declaration of ``operator()`` the input parameters
``x`` and ``y`` come first, and are passed as const pointers to arrays
of ``T``. If there were three input parameters, then the third input
parameter would come after ``y``. The output is always the last
parameter, and is also a pointer to an array. In the example above,
``e`` is a scalar, so only ``e[0]`` is set.
Then given this class definition, the auto differentiated cost
function for it can be constructed as follows.
.. code-block:: c++
CostFunction* cost_function
= new AutoDiffCostFunction<MyScalarCostFunctor, 1, 2, 2>(
new MyScalarCostFunctor(1.0)); ^ ^ ^
| | |
Dimension of residual ------+ | |
Dimension of x ----------------+ |
Dimension of y -------------------+
In this example, there is usually an instance for each measurement
of ``k``.
In the instantiation above, the template parameters following
``MyScalarCostFunction``, ``<1, 2, 2>`` describe the functor as
computing a 1-dimensional output from two arguments, both
2-dimensional.
:class:`AutoDiffCostFunction` also supports cost functions with a
runtime-determined number of residuals. For example:
.. code-block:: c++
CostFunction* cost_function
= new AutoDiffCostFunction<MyScalarCostFunctor, DYNAMIC, 2, 2>(
new CostFunctorWithDynamicNumResiduals(1.0), ^ ^ ^
runtime_number_of_residuals); <----+ | | |
| | | |
| | | |
Actual number of residuals ------+ | | |
Indicate dynamic number of residuals --------+ | |
Dimension of x ------------------------------------+ |
Dimension of y ---------------------------------------+
The framework can currently accommodate cost functions of up to 10
independent variables, and there is no limit on the dimensionality
of each of them.
**WARNING 1** A common beginner's error when first using
:class:`AutoDiffCostFunction` is to get the sizing wrong. In particular,
there is a tendency to set the template parameters to (dimension of
residual, number of parameters) instead of passing a dimension
parameter for *every parameter block*. In the example above, that
would be ``<MyScalarCostFunction, 1, 2>``, which is missing the 2
as the last template argument.
:class:`DynamicAutoDiffCostFunction`
====================================
.. class:: DynamicAutoDiffCostFunction
:class:`AutoDiffCostFunction` requires that the number of parameter
blocks and their sizes be known at compile time. It also has an
upper limit of 10 parameter blocks. In a number of applications,
this is not enough e.g., Bezier curve fitting, Neural Network
training etc.
.. code-block:: c++
template <typename CostFunctor, int Stride = 4>
class DynamicAutoDiffCostFunction : public CostFunction {
};
In such cases :class:`DynamicAutoDiffCostFunction` can be
used. Like :class:`AutoDiffCostFunction` the user must define a
templated functor, but the signature of the functor differs
slightly. The expected interface for the cost functors is:
.. code-block:: c++
struct MyCostFunctor {
template<typename T>
bool operator()(T const* const* parameters, T* residuals) const {
}
}
Since the sizing of the parameters is done at runtime, you must
also specify the sizes after creating the dynamic autodiff cost
function. For example:
.. code-block:: c++
DynamicAutoDiffCostFunction<MyCostFunctor, 4>* cost_function =
new DynamicAutoDiffCostFunction<MyCostFunctor, 4>(
new MyCostFunctor());
cost_function->AddParameterBlock(5);
cost_function->AddParameterBlock(10);
cost_function->SetNumResiduals(21);
Under the hood, the implementation evaluates the cost function
multiple times, computing a small set of the derivatives (four by
default, controlled by the ``Stride`` template parameter) with each
pass. There is a performance tradeoff with the size of the passes;
Smaller sizes are more cache efficient but result in larger number
of passes, and larger stride lengths can destroy cache-locality
while reducing the number of passes over the cost function. The
optimal value depends on the number and sizes of the various
parameter blocks.
As a rule of thumb, try using :class:`AutoDiffCostFunction` before
you use :class:`DynamicAutoDiffCostFunction`.
:class:`NumericDiffCostFunction`
================================
.. class:: NumericDiffCostFunction
In some cases, its not possible to define a templated cost functor,
for example when the evaluation of the residual involves a call to a
library function that you do not have control over. In such a
situation, `numerical differentiation
<http://en.wikipedia.org/wiki/Numerical_differentiation>`_ can be
used.
.. NOTE ::
TODO(sameeragarwal): Add documentation for the constructor and for
NumericDiffOptions. Update DynamicNumericDiffOptions in a similar
manner.
.. code-block:: c++
template <typename CostFunctor,
NumericDiffMethodType method = CENTRAL,
int kNumResiduals, // Number of residuals, or ceres::DYNAMIC.
int N0, // Number of parameters in block 0.
int N1 = 0, // Number of parameters in block 1.
int N2 = 0, // Number of parameters in block 2.
int N3 = 0, // Number of parameters in block 3.
int N4 = 0, // Number of parameters in block 4.
int N5 = 0, // Number of parameters in block 5.
int N6 = 0, // Number of parameters in block 6.
int N7 = 0, // Number of parameters in block 7.
int N8 = 0, // Number of parameters in block 8.
int N9 = 0> // Number of parameters in block 9.
class NumericDiffCostFunction : public
SizedCostFunction<kNumResiduals, N0, N1, N2, N3, N4, N5, N6, N7, N8, N9> {
};
To get a numerically differentiated :class:`CostFunction`, you must
define a class with a ``operator()`` (a functor) that computes the
residuals. The functor must write the computed value in the last
argument (the only non-``const`` one) and return ``true`` to
indicate success. Please see :class:`CostFunction` for details on
how the return value may be used to impose simple constraints on the
parameter block. e.g., an object of the form
.. code-block:: c++
struct ScalarFunctor {
public:
bool operator()(const double* const x1,
const double* const x2,
double* residuals) const;
}
For example, consider a scalar error :math:`e = k - x'y`, where both
:math:`x` and :math:`y` are two-dimensional column vector
parameters, the prime sign indicates transposition, and :math:`k` is
a constant. The form of this error, which is the difference between
a constant and an expression, is a common pattern in least squares
problems. For example, the value :math:`x'y` might be the model
expectation for a series of measurements, where there is an instance
of the cost function for each measurement :math:`k`.
To write an numerically-differentiable class:`CostFunction` for the
above model, first define the object
.. code-block:: c++
class MyScalarCostFunctor {
MyScalarCostFunctor(double k): k_(k) {}
bool operator()(const double* const x,
const double* const y,
double* residuals) const {
residuals[0] = k_ - x[0] * y[0] + x[1] * y[1];
return true;
}
private:
double k_;
};
Note that in the declaration of ``operator()`` the input parameters
``x`` and ``y`` come first, and are passed as const pointers to
arrays of ``double`` s. If there were three input parameters, then
the third input parameter would come after ``y``. The output is
always the last parameter, and is also a pointer to an array. In the
example above, the residual is a scalar, so only ``residuals[0]`` is
set.
Then given this class definition, the numerically differentiated
:class:`CostFunction` with central differences used for computing
the derivative can be constructed as follows.
.. code-block:: c++
CostFunction* cost_function
= new NumericDiffCostFunction<MyScalarCostFunctor, CENTRAL, 1, 2, 2>(
new MyScalarCostFunctor(1.0)); ^ ^ ^ ^
| | | |
Finite Differencing Scheme -+ | | |
Dimension of residual ------------+ | |
Dimension of x ----------------------+ |
Dimension of y -------------------------+
In this example, there is usually an instance for each measurement
of `k`.
In the instantiation above, the template parameters following
``MyScalarCostFunctor``, ``1, 2, 2``, describe the functor as
computing a 1-dimensional output from two arguments, both
2-dimensional.
NumericDiffCostFunction also supports cost functions with a
runtime-determined number of residuals. For example:
.. code-block:: c++
CostFunction* cost_function
= new NumericDiffCostFunction<MyScalarCostFunctor, CENTRAL, DYNAMIC, 2, 2>(
new CostFunctorWithDynamicNumResiduals(1.0), ^ ^ ^
TAKE_OWNERSHIP, | | |
runtime_number_of_residuals); <----+ | | |
| | | |
| | | |
Actual number of residuals ------+ | | |
Indicate dynamic number of residuals --------------------+ | |
Dimension of x ------------------------------------------------+ |
Dimension of y ---------------------------------------------------+
The framework can currently accommodate cost functions of up to 10
independent variables, and there is no limit on the dimensionality
of each of them.
There are three available numeric differentiation schemes in ceres-solver:
The ``FORWARD`` difference method, which approximates :math:`f'(x)`
by computing :math:`\frac{f(x+h)-f(x)}{h}`, computes the cost
function one additional time at :math:`x+h`. It is the fastest but
least accurate method.
The ``CENTRAL`` difference method is more accurate at the cost of
twice as many function evaluations than forward difference,
estimating :math:`f'(x)` by computing
:math:`\frac{f(x+h)-f(x-h)}{2h}`.
The ``RIDDERS`` difference method[Ridders]_ is an adaptive scheme
that estimates derivatives by performing multiple central
differences at varying scales. Specifically, the algorithm starts at
a certain :math:`h` and as the derivative is estimated, this step
size decreases. To conserve function evaluations and estimate the
derivative error, the method performs Richardson extrapolations
between the tested step sizes. The algorithm exhibits considerably
higher accuracy, but does so by additional evaluations of the cost
function.
Consider using ``CENTRAL`` differences to begin with. Based on the
results, either try forward difference to improve performance or
Ridders' method to improve accuracy.
**WARNING** A common beginner's error when first using
:class:`NumericDiffCostFunction` is to get the sizing wrong. In
particular, there is a tendency to set the template parameters to
(dimension of residual, number of parameters) instead of passing a
dimension parameter for *every parameter*. In the example above,
that would be ``<MyScalarCostFunctor, 1, 2>``, which is missing the
last ``2`` argument. Please be careful when setting the size
parameters.
Numeric Differentiation & LocalParameterization
-----------------------------------------------
If your cost function depends on a parameter block that must lie on
a manifold and the functor cannot be evaluated for values of that
parameter block not on the manifold then you may have problems
numerically differentiating such functors.
This is because numeric differentiation in Ceres is performed by
perturbing the individual coordinates of the parameter blocks that
a cost functor depends on. In doing so, we assume that the
parameter blocks live in an Euclidean space and ignore the
structure of manifold that they live As a result some of the
perturbations may not lie on the manifold corresponding to the
parameter block.
For example consider a four dimensional parameter block that is
interpreted as a unit Quaternion. Perturbing the coordinates of
this parameter block will violate the unit norm property of the
parameter block.
Fixing this problem requires that :class:`NumericDiffCostFunction`
be aware of the :class:`LocalParameterization` associated with each
parameter block and only generate perturbations in the local
tangent space of each parameter block.
For now this is not considered to be a serious enough problem to
warrant changing the :class:`NumericDiffCostFunction` API. Further,
in most cases it is relatively straightforward to project a point
off the manifold back onto the manifold before using it in the
functor. For example in case of the Quaternion, normalizing the
4-vector before using it does the trick.
**Alternate Interface**
For a variety of reasons, including compatibility with legacy code,
:class:`NumericDiffCostFunction` can also take
:class:`CostFunction` objects as input. The following describes
how.
To get a numerically differentiated cost function, define a
subclass of :class:`CostFunction` such that the
:func:`CostFunction::Evaluate` function ignores the ``jacobians``
parameter. The numeric differentiation wrapper will fill in the
jacobian parameter if necessary by repeatedly calling the
:func:`CostFunction::Evaluate` with small changes to the
appropriate parameters, and computing the slope. For performance,
the numeric differentiation wrapper class is templated on the
concrete cost function, even though it could be implemented only in
terms of the :class:`CostFunction` interface.
The numerically differentiated version of a cost function for a
cost function can be constructed as follows:
.. code-block:: c++
CostFunction* cost_function
= new NumericDiffCostFunction<MyCostFunction, CENTRAL, 1, 4, 8>(
new MyCostFunction(...), TAKE_OWNERSHIP);
where ``MyCostFunction`` has 1 residual and 2 parameter blocks with
sizes 4 and 8 respectively. Look at the tests for a more detailed
example.
:class:`DynamicNumericDiffCostFunction`
=======================================
.. class:: DynamicNumericDiffCostFunction
Like :class:`AutoDiffCostFunction` :class:`NumericDiffCostFunction`
requires that the number of parameter blocks and their sizes be
known at compile time. It also has an upper limit of 10 parameter
blocks. In a number of applications, this is not enough.
.. code-block:: c++
template <typename CostFunctor, NumericDiffMethodType method = CENTRAL>
class DynamicNumericDiffCostFunction : public CostFunction {
};
In such cases when numeric differentiation is desired,
:class:`DynamicNumericDiffCostFunction` can be used.
Like :class:`NumericDiffCostFunction` the user must define a
functor, but the signature of the functor differs slightly. The
expected interface for the cost functors is:
.. code-block:: c++
struct MyCostFunctor {
bool operator()(double const* const* parameters, double* residuals) const {
}
}
Since the sizing of the parameters is done at runtime, you must
also specify the sizes after creating the dynamic numeric diff cost
function. For example:
.. code-block:: c++
DynamicNumericDiffCostFunction<MyCostFunctor>* cost_function =
new DynamicNumericDiffCostFunction<MyCostFunctor>(new MyCostFunctor);
cost_function->AddParameterBlock(5);
cost_function->AddParameterBlock(10);
cost_function->SetNumResiduals(21);
As a rule of thumb, try using :class:`NumericDiffCostFunction` before
you use :class:`DynamicNumericDiffCostFunction`.
**WARNING** The same caution about mixing local parameterizations
with numeric differentiation applies as is the case with
:class:`NumericDiffCostFunction`.
:class:`CostFunctionToFunctor`
==============================
.. class:: CostFunctionToFunctor
:class:`CostFunctionToFunctor` is an adapter class that allows
users to use :class:`CostFunction` objects in templated functors
which are to be used for automatic differentiation. This allows
the user to seamlessly mix analytic, numeric and automatic
differentiation.
For example, let us assume that
.. code-block:: c++
class IntrinsicProjection : public SizedCostFunction<2, 5, 3> {
public:
IntrinsicProjection(const double* observation);
virtual bool Evaluate(double const* const* parameters,
double* residuals,
double** jacobians) const;
};
is a :class:`CostFunction` that implements the projection of a
point in its local coordinate system onto its image plane and
subtracts it from the observed point projection. It can compute its
residual and either via analytic or numerical differentiation can
compute its jacobians.
Now we would like to compose the action of this
:class:`CostFunction` with the action of camera extrinsics, i.e.,
rotation and translation. Say we have a templated function
.. code-block:: c++
template<typename T>
void RotateAndTranslatePoint(const T* rotation,
const T* translation,
const T* point,
T* result);
Then we can now do the following,
.. code-block:: c++
struct CameraProjection {
CameraProjection(double* observation)
: intrinsic_projection_(new IntrinsicProjection(observation)) {
}
template <typename T>
bool operator()(const T* rotation,
const T* translation,
const T* intrinsics,
const T* point,
T* residual) const {
T transformed_point[3];
RotateAndTranslatePoint(rotation, translation, point, transformed_point);
// Note that we call intrinsic_projection_, just like it was
// any other templated functor.
return intrinsic_projection_(intrinsics, transformed_point, residual);
}
private:
CostFunctionToFunctor<2,5,3> intrinsic_projection_;
};
Note that :class:`CostFunctionToFunctor` takes ownership of the
:class:`CostFunction` that was passed in to the constructor.
In the above example, we assumed that ``IntrinsicProjection`` is a
``CostFunction`` capable of evaluating its value and its
derivatives. Suppose, if that were not the case and
``IntrinsicProjection`` was defined as follows:
.. code-block:: c++
struct IntrinsicProjection {
IntrinsicProjection(const double* observation) {
observation_[0] = observation[0];
observation_[1] = observation[1];
}
bool operator()(const double* calibration,
const double* point,
double* residuals) const {
double projection[2];
ThirdPartyProjectionFunction(calibration, point, projection);
residuals[0] = observation_[0] - projection[0];
residuals[1] = observation_[1] - projection[1];
return true;
}
double observation_[2];
};
Here ``ThirdPartyProjectionFunction`` is some third party library
function that we have no control over. So this function can compute
its value and we would like to use numeric differentiation to
compute its derivatives. In this case we can use a combination of
``NumericDiffCostFunction`` and ``CostFunctionToFunctor`` to get the
job done.
.. code-block:: c++
struct CameraProjection {
CameraProjection(double* observation)
: intrinsic_projection_(
new NumericDiffCostFunction<IntrinsicProjection, CENTRAL, 2, 5, 3>(
new IntrinsicProjection(observation))) {}
template <typename T>
bool operator()(const T* rotation,
const T* translation,
const T* intrinsics,
const T* point,
T* residuals) const {
T transformed_point[3];
RotateAndTranslatePoint(rotation, translation, point, transformed_point);
return intrinsic_projection_(intrinsics, transformed_point, residuals);
}
private:
CostFunctionToFunctor<2, 5, 3> intrinsic_projection_;
};
:class:`DynamicCostFunctionToFunctor`
=====================================
.. class:: DynamicCostFunctionToFunctor
:class:`DynamicCostFunctionToFunctor` provides the same functionality as
:class:`CostFunctionToFunctor` for cases where the number and size of the
parameter vectors and residuals are not known at compile-time. The API
provided by :class:`DynamicCostFunctionToFunctor` matches what would be
expected by :class:`DynamicAutoDiffCostFunction`, i.e. it provides a
templated functor of this form:
.. code-block:: c++
template<typename T>
bool operator()(T const* const* parameters, T* residuals) const;
Similar to the example given for :class:`CostFunctionToFunctor`, let us
assume that
.. code-block:: c++
class IntrinsicProjection : public CostFunction {
public:
IntrinsicProjection(const double* observation);
virtual bool Evaluate(double const* const* parameters,
double* residuals,
double** jacobians) const;
};
is a :class:`CostFunction` that projects a point in its local coordinate
system onto its image plane and subtracts it from the observed point
projection.
Using this :class:`CostFunction` in a templated functor would then look like
this:
.. code-block:: c++
struct CameraProjection {
CameraProjection(double* observation)
: intrinsic_projection_(new IntrinsicProjection(observation)) {
}
template <typename T>
bool operator()(T const* const* parameters,
T* residual) const {
const T* rotation = parameters[0];
const T* translation = parameters[1];
const T* intrinsics = parameters[2];
const T* point = parameters[3];
T transformed_point[3];
RotateAndTranslatePoint(rotation, translation, point, transformed_point);
const T* projection_parameters[2];
projection_parameters[0] = intrinsics;
projection_parameters[1] = transformed_point;
return intrinsic_projection_(projection_parameters, residual);
}
private:
DynamicCostFunctionToFunctor intrinsic_projection_;
};
Like :class:`CostFunctionToFunctor`, :class:`DynamicCostFunctionToFunctor`
takes ownership of the :class:`CostFunction` that was passed in to the
constructor.
:class:`ConditionedCostFunction`
================================
.. class:: ConditionedCostFunction
This class allows you to apply different conditioning to the residual
values of a wrapped cost function. An example where this is useful is
where you have an existing cost function that produces N values, but you
want the total cost to be something other than just the sum of these
squared values - maybe you want to apply a different scaling to some
values, to change their contribution to the cost.
Usage:
.. code-block:: c++
// my_cost_function produces N residuals
CostFunction* my_cost_function = ...
CHECK_EQ(N, my_cost_function->num_residuals());
vector<CostFunction*> conditioners;
// Make N 1x1 cost functions (1 parameter, 1 residual)
CostFunction* f_1 = ...
conditioners.push_back(f_1);
CostFunction* f_N = ...
conditioners.push_back(f_N);
ConditionedCostFunction* ccf =
new ConditionedCostFunction(my_cost_function, conditioners);
Now ``ccf`` 's ``residual[i]`` (i=0..N-1) will be passed though the
:math:`i^{\text{th}}` conditioner.
.. code-block:: c++
ccf_residual[i] = f_i(my_cost_function_residual[i])
and the Jacobian will be affected appropriately.
:class:`GradientChecker`
================================
.. class:: GradientChecker
This class compares the Jacobians returned by a cost function against
derivatives estimated using finite differencing. It is meant as a tool for
unit testing, giving you more fine-grained control than the check_gradients
option in the solver options.
The condition enforced is that
.. math:: \forall{i,j}: \frac{J_{ij} - J'_{ij}}{max_{ij}(J_{ij} - J'_{ij})} < r
where :math:`J_{ij}` is the jacobian as computed by the supplied cost
function (by the user) multiplied by the local parameterization Jacobian,
:math:`J'_{ij}` is the jacobian as computed by finite differences,
multiplied by the local parameterization Jacobian as well, and :math:`r`
is the relative precision.
Usage:
.. code-block:: c++
// my_cost_function takes two parameter blocks. The first has a local
// parameterization associated with it.
CostFunction* my_cost_function = ...
LocalParameterization* my_parameterization = ...
NumericDiffOptions numeric_diff_options;
std::vector<LocalParameterization*> local_parameterizations;
local_parameterizations.push_back(my_parameterization);
local_parameterizations.push_back(NULL);
std::vector parameter1;
std::vector parameter2;
// Fill parameter 1 & 2 with test data...
std::vector<double*> parameter_blocks;
parameter_blocks.push_back(parameter1.data());
parameter_blocks.push_back(parameter2.data());
GradientChecker gradient_checker(my_cost_function,
local_parameterizations, numeric_diff_options);
GradientCheckResults results;
if (!gradient_checker.Probe(parameter_blocks.data(), 1e-9, &results) {
LOG(ERROR) << "An error has occurred:\n" << results.error_log;
}
:class:`NormalPrior`
====================
.. class:: NormalPrior
.. code-block:: c++
class NormalPrior: public CostFunction {
public:
// Check that the number of rows in the vector b are the same as the
// number of columns in the matrix A, crash otherwise.
NormalPrior(const Matrix& A, const Vector& b);
virtual bool Evaluate(double const* const* parameters,
double* residuals,
double** jacobians) const;
};
Implements a cost function of the form
.. math:: cost(x) = ||A(x - b)||^2
where, the matrix :math:`A` and the vector :math:`b` are fixed and :math:`x`
is the variable. In case the user is interested in implementing a cost
function of the form
.. math:: cost(x) = (x - \mu)^T S^{-1} (x - \mu)
where, :math:`\mu` is a vector and :math:`S` is a covariance matrix,
then, :math:`A = S^{-1/2}`, i.e the matrix :math:`A` is the square
root of the inverse of the covariance, also known as the stiffness
matrix. There are however no restrictions on the shape of
:math:`A`. It is free to be rectangular, which would be the case if
the covariance matrix :math:`S` is rank deficient.
.. _`section-loss_function`:
:class:`LossFunction`
=====================
.. class:: LossFunction
For least squares problems where the minimization may encounter
input terms that contain outliers, that is, completely bogus
measurements, it is important to use a loss function that reduces
their influence.
Consider a structure from motion problem. The unknowns are 3D
points and camera parameters, and the measurements are image
coordinates describing the expected reprojected position for a
point in a camera. For example, we want to model the geometry of a
street scene with fire hydrants and cars, observed by a moving
camera with unknown parameters, and the only 3D points we care
about are the pointy tippy-tops of the fire hydrants. Our magic
image processing algorithm, which is responsible for producing the
measurements that are input to Ceres, has found and matched all
such tippy-tops in all image frames, except that in one of the
frame it mistook a car's headlight for a hydrant. If we didn't do
anything special the residual for the erroneous measurement will
result in the entire solution getting pulled away from the optimum
to reduce the large error that would otherwise be attributed to the
wrong measurement.
Using a robust loss function, the cost for large residuals is
reduced. In the example above, this leads to outlier terms getting
down-weighted so they do not overly influence the final solution.
.. code-block:: c++
class LossFunction {
public:
virtual void Evaluate(double s, double out[3]) const = 0;
};
The key method is :func:`LossFunction::Evaluate`, which given a
non-negative scalar ``s``, computes
.. math:: out = \begin{bmatrix}\rho(s), & \rho'(s), & \rho''(s)\end{bmatrix}
Here the convention is that the contribution of a term to the cost
function is given by :math:`\frac{1}{2}\rho(s)`, where :math:`s
=\|f_i\|^2`. Calling the method with a negative value of :math:`s`
is an error and the implementations are not required to handle that
case.
Most sane choices of :math:`\rho` satisfy:
.. math::
\rho(0) &= 0\\
\rho'(0) &= 1\\
\rho'(s) &< 1 \text{ in the outlier region}\\
\rho''(s) &< 0 \text{ in the outlier region}
so that they mimic the squared cost for small residuals.
**Scaling**
Given one robustifier :math:`\rho(s)` one can change the length
scale at which robustification takes place, by adding a scale
factor :math:`a > 0` which gives us :math:`\rho(s,a) = a^2 \rho(s /
a^2)` and the first and second derivatives as :math:`\rho'(s /
a^2)` and :math:`(1 / a^2) \rho''(s / a^2)` respectively.
The reason for the appearance of squaring is that :math:`a` is in
the units of the residual vector norm whereas :math:`s` is a squared
norm. For applications it is more convenient to specify :math:`a` than
its square.
Instances
---------
Ceres includes a number of predefined loss functions. For simplicity
we described their unscaled versions. The figure below illustrates
their shape graphically. More details can be found in
``include/ceres/loss_function.h``.
.. figure:: loss.png
:figwidth: 500px
:height: 400px
:align: center
Shape of the various common loss functions.
.. class:: TrivialLoss
.. math:: \rho(s) = s
.. class:: HuberLoss
.. math:: \rho(s) = \begin{cases} s & s \le 1\\ 2 \sqrt{s} - 1 & s > 1 \end{cases}
.. class:: SoftLOneLoss
.. math:: \rho(s) = 2 (\sqrt{1+s} - 1)
.. class:: CauchyLoss
.. math:: \rho(s) = \log(1 + s)
.. class:: ArctanLoss
.. math:: \rho(s) = \arctan(s)
.. class:: TolerantLoss
.. math:: \rho(s,a,b) = b \log(1 + e^{(s - a) / b}) - b \log(1 + e^{-a / b})
.. class:: ComposedLoss
Given two loss functions ``f`` and ``g``, implements the loss
function ``h(s) = f(g(s))``.
.. code-block:: c++
class ComposedLoss : public LossFunction {
public:
explicit ComposedLoss(const LossFunction* f,
Ownership ownership_f,
const LossFunction* g,
Ownership ownership_g);
};
.. class:: ScaledLoss
Sometimes you want to simply scale the output value of the
robustifier. For example, you might want to weight different error
terms differently (e.g., weight pixel reprojection errors
differently from terrain errors).
Given a loss function :math:`\rho(s)` and a scalar :math:`a`, :class:`ScaledLoss`
implements the function :math:`a \rho(s)`.
Since we treat a ``NULL`` Loss function as the Identity loss
function, :math:`rho` = ``NULL``: is a valid input and will result
in the input being scaled by :math:`a`. This provides a simple way
of implementing a scaled ResidualBlock.
.. class:: LossFunctionWrapper
Sometimes after the optimization problem has been constructed, we
wish to mutate the scale of the loss function. For example, when
performing estimation from data which has substantial outliers,
convergence can be improved by starting out with a large scale,
optimizing the problem and then reducing the scale. This can have
better convergence behavior than just using a loss function with a
small scale.
This templated class allows the user to implement a loss function
whose scale can be mutated after an optimization problem has been
constructed, e.g,
.. code-block:: c++
Problem problem;
// Add parameter blocks
CostFunction* cost_function =
new AutoDiffCostFunction < UW_Camera_Mapper, 2, 9, 3>(
new UW_Camera_Mapper(feature_x, feature_y));
LossFunctionWrapper* loss_function(new HuberLoss(1.0), TAKE_OWNERSHIP);
problem.AddResidualBlock(cost_function, loss_function, parameters);
Solver::Options options;
Solver::Summary summary;
Solve(options, &problem, &summary);
loss_function->Reset(new HuberLoss(1.0), TAKE_OWNERSHIP);
Solve(options, &problem, &summary);
Theory
------
Let us consider a problem with a single problem and a single parameter
block.
.. math::
\min_x \frac{1}{2}\rho(f^2(x))
Then, the robustified gradient and the Gauss-Newton Hessian are
.. math::
g(x) &= \rho'J^\top(x)f(x)\\
H(x) &= J^\top(x)\left(\rho' + 2 \rho''f(x)f^\top(x)\right)J(x)
where the terms involving the second derivatives of :math:`f(x)` have
been ignored. Note that :math:`H(x)` is indefinite if
:math:`\rho''f(x)^\top f(x) + \frac{1}{2}\rho' < 0`. If this is not
the case, then its possible to re-weight the residual and the Jacobian
matrix such that the corresponding linear least squares problem for
the robustified Gauss-Newton step.
Let :math:`\alpha` be a root of
.. math:: \frac{1}{2}\alpha^2 - \alpha - \frac{\rho''}{\rho'}\|f(x)\|^2 = 0.
Then, define the rescaled residual and Jacobian as
.. math::
\tilde{f}(x) &= \frac{\sqrt{\rho'}}{1 - \alpha} f(x)\\
\tilde{J}(x) &= \sqrt{\rho'}\left(1 - \alpha
\frac{f(x)f^\top(x)}{\left\|f(x)\right\|^2} \right)J(x)
In the case :math:`2 \rho''\left\|f(x)\right\|^2 + \rho' \lesssim 0`,
we limit :math:`\alpha \le 1- \epsilon` for some small
:math:`\epsilon`. For more details see [Triggs]_.
With this simple rescaling, one can use any Jacobian based non-linear
least squares algorithm to robustified non-linear least squares
problems.
:class:`LocalParameterization`
==============================
.. class:: LocalParameterization
.. code-block:: c++
class LocalParameterization {
public:
virtual ~LocalParameterization() {}
virtual bool Plus(const double* x,
const double* delta,
double* x_plus_delta) const = 0;
virtual bool ComputeJacobian(const double* x, double* jacobian) const = 0;
virtual bool MultiplyByJacobian(const double* x,
const int num_rows,
const double* global_matrix,
double* local_matrix) const;
virtual int GlobalSize() const = 0;
virtual int LocalSize() const = 0;
};
Sometimes the parameters :math:`x` can overparameterize a
problem. In that case it is desirable to choose a parameterization
to remove the null directions of the cost. More generally, if
:math:`x` lies on a manifold of a smaller dimension than the
ambient space that it is embedded in, then it is numerically and
computationally more effective to optimize it using a
parameterization that lives in the tangent space of that manifold
at each point.
For example, a sphere in three dimensions is a two dimensional
manifold, embedded in a three dimensional space. At each point on
the sphere, the plane tangent to it defines a two dimensional
tangent space. For a cost function defined on this sphere, given a
point :math:`x`, moving in the direction normal to the sphere at
that point is not useful. Thus a better way to parameterize a point
on a sphere is to optimize over two dimensional vector
:math:`\Delta x` in the tangent space at the point on the sphere
point and then "move" to the point :math:`x + \Delta x`, where the
move operation involves projecting back onto the sphere. Doing so
removes a redundant dimension from the optimization, making it
numerically more robust and efficient.
More generally we can define a function
.. math:: x' = \boxplus(x, \Delta x),
where :math:`x'` has the same size as :math:`x`, and :math:`\Delta
x` is of size less than or equal to :math:`x`. The function
:math:`\boxplus`, generalizes the definition of vector
addition. Thus it satisfies the identity
.. math:: \boxplus(x, 0) = x,\quad \forall x.
Instances of :class:`LocalParameterization` implement the
:math:`\boxplus` operation and its derivative with respect to
:math:`\Delta x` at :math:`\Delta x = 0`.
.. function:: int LocalParameterization::GlobalSize()
The dimension of the ambient space in which the parameter block
:math:`x` lives.
.. function:: int LocalParameterization::LocalSize()
The size of the tangent space
that :math:`\Delta x` lives in.
.. function:: bool LocalParameterization::Plus(const double* x, const double* delta, double* x_plus_delta) const
:func:`LocalParameterization::Plus` implements :math:`\boxplus(x,\Delta x)`.
.. function:: bool LocalParameterization::ComputeJacobian(const double* x, double* jacobian) const
Computes the Jacobian matrix
.. math:: J = \left . \frac{\partial }{\partial \Delta x} \boxplus(x,\Delta x)\right|_{\Delta x = 0}
in row major form.
.. function:: bool MultiplyByJacobian(const double* x, const int num_rows, const double* global_matrix, double* local_matrix) const
local_matrix = global_matrix * jacobian
global_matrix is a num_rows x GlobalSize row major matrix.
local_matrix is a num_rows x LocalSize row major matrix.
jacobian is the matrix returned by :func:`LocalParameterization::ComputeJacobian` at :math:`x`.
This is only used by GradientProblem. For most normal uses, it is
okay to use the default implementation.
Instances
---------
.. class:: IdentityParameterization
A trivial version of :math:`\boxplus` is when :math:`\Delta x` is
of the same size as :math:`x` and
.. math:: \boxplus(x, \Delta x) = x + \Delta x
.. class:: SubsetParameterization
A more interesting case if :math:`x` is a two dimensional vector,
and the user wishes to hold the first coordinate constant. Then,
:math:`\Delta x` is a scalar and :math:`\boxplus` is defined as
.. math::
\boxplus(x, \Delta x) = x + \left[ \begin{array}{c} 0 \\ 1
\end{array} \right] \Delta x
:class:`SubsetParameterization` generalizes this construction to
hold any part of a parameter block constant.
.. class:: QuaternionParameterization
Another example that occurs commonly in Structure from Motion
problems is when camera rotations are parameterized using a
quaternion. There, it is useful only to make updates orthogonal to
that 4-vector defining the quaternion. One way to do this is to let
:math:`\Delta x` be a 3 dimensional vector and define
:math:`\boxplus` to be
.. math:: \boxplus(x, \Delta x) = \left[ \cos(|\Delta x|), \frac{\sin\left(|\Delta x|\right)}{|\Delta x|} \Delta x \right] * x
:label: quaternion
The multiplication between the two 4-vectors on the right hand side
is the standard quaternion
product. :class:`QuaternionParameterization` is an implementation
of :eq:`quaternion`.
.. class:: EigenQuaternionParameterization
Eigen uses a different internal memory layout for the elements of the
quaternion than what is commonly used. Specifically, Eigen stores the
elements in memory as [x, y, z, w] where the real part is last
whereas it is typically stored first. Note, when creating an Eigen
quaternion through the constructor the elements are accepted in w, x,
y, z order. Since Ceres operates on parameter blocks which are raw
double pointers this difference is important and requires a different
parameterization. :class:`EigenQuaternionParameterization` uses the
same update as :class:`QuaternionParameterization` but takes into
account Eigen's internal memory element ordering.
.. class:: HomogeneousVectorParameterization
In computer vision, homogeneous vectors are commonly used to
represent entities in projective geometry such as points in
projective space. One example where it is useful to use this
over-parameterization is in representing points whose triangulation
is ill-conditioned. Here it is advantageous to use homogeneous
vectors, instead of an Euclidean vector, because it can represent
points at infinity.
When using homogeneous vectors it is useful to only make updates
orthogonal to that :math:`n`-vector defining the homogeneous
vector [HartleyZisserman]_. One way to do this is to let :math:`\Delta x`
be a :math:`n-1` dimensional vector and define :math:`\boxplus` to be
.. math:: \boxplus(x, \Delta x) = \left[ \frac{\sin\left(0.5 |\Delta x|\right)}{|\Delta x|} \Delta x, \cos(0.5 |\Delta x|) \right] * x
The multiplication between the two vectors on the right hand side
is defined as an operator which applies the update orthogonal to
:math:`x` to remain on the sphere. Note, it is assumed that
last element of :math:`x` is the scalar component of the homogeneous
vector.
.. class:: ProductParameterization
Consider an optimization problem over the space of rigid
transformations :math:`SE(3)`, which is the Cartesian product of
:math:`SO(3)` and :math:`\mathbb{R}^3`. Suppose you are using
Quaternions to represent the rotation, Ceres ships with a local
parameterization for that and :math:`\mathbb{R}^3` requires no, or
:class:`IdentityParameterization` parameterization. So how do we
construct a local parameterization for a parameter block a rigid
transformation?
In cases, where a parameter block is the Cartesian product of a
number of manifolds and you have the local parameterization of the
individual manifolds available, :class:`ProductParameterization`
can be used to construct a local parameterization of the cartesian
product. For the case of the rigid transformation, where say you
have a parameter block of size 7, where the first four entries
represent the rotation as a quaternion, a local parameterization
can be constructed as
.. code-block:: c++
ProductParameterization se3_param(new QuaternionParameterization(),
new IdentityParameterization(3));
:class:`AutoDiffLocalParameterization`
======================================
.. class:: AutoDiffLocalParameterization
:class:`AutoDiffLocalParameterization` does for
:class:`LocalParameterization` what :class:`AutoDiffCostFunction`
does for :class:`CostFunction`. It allows the user to define a
templated functor that implements the
:func:`LocalParameterization::Plus` operation and it uses automatic
differentiation to implement the computation of the Jacobian.
To get an auto differentiated local parameterization, you must
define a class with a templated operator() (a functor) that computes
.. math:: x' = \boxplus(x, \Delta x),
For example, Quaternions have a three dimensional local
parameterization. Its plus operation can be implemented as (taken
from `internal/ceres/autodiff_local_parameterization_test.cc
<https://ceres-solver.googlesource.com/ceres-solver/+/master/internal/ceres/autodiff_local_parameterization_test.cc>`_
)
.. code-block:: c++
struct QuaternionPlus {
template<typename T>
bool operator()(const T* x, const T* delta, T* x_plus_delta) const {
const T squared_norm_delta =
delta[0] * delta[0] + delta[1] * delta[1] + delta[2] * delta[2];
T q_delta[4];
if (squared_norm_delta > 0.0) {
T norm_delta = sqrt(squared_norm_delta);
const T sin_delta_by_delta = sin(norm_delta) / norm_delta;
q_delta[0] = cos(norm_delta);
q_delta[1] = sin_delta_by_delta * delta[0];
q_delta[2] = sin_delta_by_delta * delta[1];
q_delta[3] = sin_delta_by_delta * delta[2];
} else {
// We do not just use q_delta = [1,0,0,0] here because that is a
// constant and when used for automatic differentiation will
// lead to a zero derivative. Instead we take a first order
// approximation and evaluate it at zero.
q_delta[0] = T(1.0);
q_delta[1] = delta[0];
q_delta[2] = delta[1];
q_delta[3] = delta[2];
}
Quaternionproduct(q_delta, x, x_plus_delta);
return true;
}
};
Given this struct, the auto differentiated local
parameterization can now be constructed as
.. code-block:: c++
LocalParameterization* local_parameterization =
new AutoDiffLocalParameterization<QuaternionPlus, 4, 3>;
| |
Global Size ---------------+ |
Local Size -------------------+
:class:`Problem`
================
.. class:: Problem
:class:`Problem` holds the robustified bounds constrained
non-linear least squares problem :eq:`ceresproblem_modeling`. To
create a least squares problem, use the
:func:`Problem::AddResidualBlock` and
:func:`Problem::AddParameterBlock` methods.
For example a problem containing 3 parameter blocks of sizes 3, 4
and 5 respectively and two residual blocks of size 2 and 6:
.. code-block:: c++
double x1[] = { 1.0, 2.0, 3.0 };
double x2[] = { 1.0, 2.0, 3.0, 5.0 };
double x3[] = { 1.0, 2.0, 3.0, 6.0, 7.0 };
Problem problem;
problem.AddResidualBlock(new MyUnaryCostFunction(...), x1);
problem.AddResidualBlock(new MyBinaryCostFunction(...), x2, x3);
:func:`Problem::AddResidualBlock` as the name implies, adds a
residual block to the problem. It adds a :class:`CostFunction`, an
optional :class:`LossFunction` and connects the
:class:`CostFunction` to a set of parameter block.
The cost function carries with it information about the sizes of
the parameter blocks it expects. The function checks that these
match the sizes of the parameter blocks listed in
``parameter_blocks``. The program aborts if a mismatch is
detected. ``loss_function`` can be ``NULL``, in which case the cost
of the term is just the squared norm of the residuals.
The user has the option of explicitly adding the parameter blocks
using :func:`Problem::AddParameterBlock`. This causes additional
correctness checking; however, :func:`Problem::AddResidualBlock`
implicitly adds the parameter blocks if they are not present, so
calling :func:`Problem::AddParameterBlock` explicitly is not
required.
:func:`Problem::AddParameterBlock` explicitly adds a parameter
block to the :class:`Problem`. Optionally it allows the user to
associate a :class:`LocalParameterization` object with the
parameter block too. Repeated calls with the same arguments are
ignored. Repeated calls with the same double pointer but a
different size results in undefined behavior.
You can set any parameter block to be constant using
:func:`Problem::SetParameterBlockConstant` and undo this using
:func:`SetParameterBlockVariable`.
In fact you can set any number of parameter blocks to be constant,
and Ceres is smart enough to figure out what part of the problem
you have constructed depends on the parameter blocks that are free
to change and only spends time solving it. So for example if you
constructed a problem with a million parameter blocks and 2 million
residual blocks, but then set all but one parameter blocks to be
constant and say only 10 residual blocks depend on this one
non-constant parameter block. Then the computational effort Ceres
spends in solving this problem will be the same if you had defined
a problem with one parameter block and 10 residual blocks.
**Ownership**
:class:`Problem` by default takes ownership of the
``cost_function``, ``loss_function`` and ``local_parameterization``
pointers. These objects remain live for the life of the
:class:`Problem`. If the user wishes to keep control over the
destruction of these objects, then they can do this by setting the
corresponding enums in the :class:`Problem::Options` struct.
Note that even though the Problem takes ownership of ``cost_function``
and ``loss_function``, it does not preclude the user from re-using
them in another residual block. The destructor takes care to call
delete on each ``cost_function`` or ``loss_function`` pointer only
once, regardless of how many residual blocks refer to them.
.. function:: ResidualBlockId Problem::AddResidualBlock(CostFunction* cost_function, LossFunction* loss_function, const vector<double*> parameter_blocks)
.. function:: ResidualBlockId Problem::AddResidualBlock(CostFunction* cost_function, LossFunction* loss_function, double *x0, double *x1, ...)
Add a residual block to the overall cost function. The cost
function carries with it information about the sizes of the
parameter blocks it expects. The function checks that these match
the sizes of the parameter blocks listed in parameter_blocks. The
program aborts if a mismatch is detected. loss_function can be
NULL, in which case the cost of the term is just the squared norm
of the residuals.
The parameter blocks may be passed together as a
``vector<double*>``, or as up to ten separate ``double*`` pointers.
The user has the option of explicitly adding the parameter blocks
using AddParameterBlock. This causes additional correctness
checking; however, AddResidualBlock implicitly adds the parameter
blocks if they are not present, so calling AddParameterBlock
explicitly is not required.
The Problem object by default takes ownership of the
cost_function and loss_function pointers. These objects remain
live for the life of the Problem object. If the user wishes to
keep control over the destruction of these objects, then they can
do this by setting the corresponding enums in the Options struct.
Note: Even though the Problem takes ownership of cost_function
and loss_function, it does not preclude the user from re-using
them in another residual block. The destructor takes care to call
delete on each cost_function or loss_function pointer only once,
regardless of how many residual blocks refer to them.
Example usage:
.. code-block:: c++
double x1[] = {1.0, 2.0, 3.0};
double x2[] = {1.0, 2.0, 5.0, 6.0};
double x3[] = {3.0, 6.0, 2.0, 5.0, 1.0};
vector<double*> v1;
v1.push_back(x1);
vector<double*> v2;
v2.push_back(x2);
v2.push_back(x1);
Problem problem;
problem.AddResidualBlock(new MyUnaryCostFunction(...), NULL, x1);
problem.AddResidualBlock(new MyBinaryCostFunction(...), NULL, x2, x1);
problem.AddResidualBlock(new MyUnaryCostFunction(...), NULL, v1);
problem.AddResidualBlock(new MyBinaryCostFunction(...), NULL, v2);
.. function:: void Problem::AddParameterBlock(double* values, int size, LocalParameterization* local_parameterization)
Add a parameter block with appropriate size to the problem.
Repeated calls with the same arguments are ignored. Repeated calls
with the same double pointer but a different size results in
undefined behavior.
.. function:: void Problem::AddParameterBlock(double* values, int size)
Add a parameter block with appropriate size and parameterization to
the problem. Repeated calls with the same arguments are
ignored. Repeated calls with the same double pointer but a
different size results in undefined behavior.
.. function:: void Problem::RemoveResidualBlock(ResidualBlockId residual_block)
Remove a residual block from the problem. Any parameters that the residual
block depends on are not removed. The cost and loss functions for the
residual block will not get deleted immediately; won't happen until the
problem itself is deleted. If Problem::Options::enable_fast_removal is
true, then the removal is fast (almost constant time). Otherwise, removing a
residual block will incur a scan of the entire Problem object to verify that
the residual_block represents a valid residual in the problem.
**WARNING:** Removing a residual or parameter block will destroy
the implicit ordering, rendering the jacobian or residuals returned
from the solver uninterpretable. If you depend on the evaluated
jacobian, do not use remove! This may change in a future release.
Hold the indicated parameter block constant during optimization.
.. function:: void Problem::RemoveParameterBlock(double* values)
Remove a parameter block from the problem. The parameterization of
the parameter block, if it exists, will persist until the deletion
of the problem (similar to cost/loss functions in residual block
removal). Any residual blocks that depend on the parameter are also
removed, as described above in RemoveResidualBlock(). If
Problem::Options::enable_fast_removal is true, then
the removal is fast (almost constant time). Otherwise, removing a
parameter block will incur a scan of the entire Problem object.
**WARNING:** Removing a residual or parameter block will destroy
the implicit ordering, rendering the jacobian or residuals returned
from the solver uninterpretable. If you depend on the evaluated
jacobian, do not use remove! This may change in a future release.
.. function:: void Problem::SetParameterBlockConstant(double* values)
Hold the indicated parameter block constant during optimization.
.. function:: void Problem::SetParameterBlockVariable(double* values)
Allow the indicated parameter to vary during optimization.
.. function:: void Problem::SetParameterization(double* values, LocalParameterization* local_parameterization)
Set the local parameterization for one of the parameter blocks.
The local_parameterization is owned by the Problem by default. It
is acceptable to set the same parameterization for multiple
parameters; the destructor is careful to delete local
parameterizations only once. The local parameterization can only be
set once per parameter, and cannot be changed once set.
.. function:: LocalParameterization* Problem::GetParameterization(double* values) const
Get the local parameterization object associated with this
parameter block. If there is no parameterization object associated
then `NULL` is returned
.. function:: void Problem::SetParameterLowerBound(double* values, int index, double lower_bound)
Set the lower bound for the parameter at position `index` in the
parameter block corresponding to `values`. By default the lower
bound is ``-std::numeric_limits<double>::max()``, which is treated
by the solver as the same as :math:`-\infty`.
.. function:: void Problem::SetParameterUpperBound(double* values, int index, double upper_bound)
Set the upper bound for the parameter at position `index` in the
parameter block corresponding to `values`. By default the value is
``std::numeric_limits<double>::max()``, which is treated by the
solver as the same as :math:`\infty`.
.. function:: double Problem::GetParameterLowerBound(double* values, int index)
Get the lower bound for the parameter with position `index`. If the
parameter is not bounded by the user, then its lower bound is
``-std::numeric_limits<double>::max()``.
.. function:: double Problem::GetParameterUpperBound(double* values, int index)
Get the upper bound for the parameter with position `index`. If the
parameter is not bounded by the user, then its upper bound is
``std::numeric_limits<double>::max()``.
.. function:: int Problem::NumParameterBlocks() const
Number of parameter blocks in the problem. Always equals
parameter_blocks().size() and parameter_block_sizes().size().
.. function:: int Problem::NumParameters() const
The size of the parameter vector obtained by summing over the sizes
of all the parameter blocks.
.. function:: int Problem::NumResidualBlocks() const
Number of residual blocks in the problem. Always equals
residual_blocks().size().
.. function:: int Problem::NumResiduals() const
The size of the residual vector obtained by summing over the sizes
of all of the residual blocks.
.. function:: int Problem::ParameterBlockSize(const double* values) const
The size of the parameter block.
.. function:: int Problem::ParameterBlockLocalSize(const double* values) const
The size of local parameterization for the parameter block. If
there is no local parameterization associated with this parameter
block, then ``ParameterBlockLocalSize`` = ``ParameterBlockSize``.
.. function:: bool Problem::HasParameterBlock(const double* values) const
Is the given parameter block present in the problem or not?
.. function:: void Problem::GetParameterBlocks(vector<double*>* parameter_blocks) const
Fills the passed ``parameter_blocks`` vector with pointers to the
parameter blocks currently in the problem. After this call,
``parameter_block.size() == NumParameterBlocks``.
.. function:: void Problem::GetResidualBlocks(vector<ResidualBlockId>* residual_blocks) const
Fills the passed `residual_blocks` vector with pointers to the
residual blocks currently in the problem. After this call,
`residual_blocks.size() == NumResidualBlocks`.
.. function:: void Problem::GetParameterBlocksForResidualBlock(const ResidualBlockId residual_block, vector<double*>* parameter_blocks) const
Get all the parameter blocks that depend on the given residual
block.
.. function:: void Problem::GetResidualBlocksForParameterBlock(const double* values, vector<ResidualBlockId>* residual_blocks) const
Get all the residual blocks that depend on the given parameter
block.
If `Problem::Options::enable_fast_removal` is
`true`, then getting the residual blocks is fast and depends only
on the number of residual blocks. Otherwise, getting the residual
blocks for a parameter block will incur a scan of the entire
:class:`Problem` object.
.. function:: const CostFunction* GetCostFunctionForResidualBlock(const ResidualBlockId residual_block) const
Get the :class:`CostFunction` for the given residual block.
.. function:: const LossFunction* GetLossFunctionForResidualBlock(const ResidualBlockId residual_block) const
Get the :class:`LossFunction` for the given residual block.
.. function:: bool Problem::Evaluate(const Problem::EvaluateOptions& options, double* cost, vector<double>* residuals, vector<double>* gradient, CRSMatrix* jacobian)
Evaluate a :class:`Problem`. Any of the output pointers can be
`NULL`. Which residual blocks and parameter blocks are used is
controlled by the :class:`Problem::EvaluateOptions` struct below.
.. NOTE::
The evaluation will use the values stored in the memory
locations pointed to by the parameter block pointers used at the
time of the construction of the problem, for example in the
following code:
.. code-block:: c++
Problem problem;
double x = 1;
problem.Add(new MyCostFunction, NULL, &x);
double cost = 0.0;
problem.Evaluate(Problem::EvaluateOptions(), &cost, NULL, NULL, NULL);
The cost is evaluated at `x = 1`. If you wish to evaluate the
problem at `x = 2`, then
.. code-block:: c++
x = 2;
problem.Evaluate(Problem::EvaluateOptions(), &cost, NULL, NULL, NULL);
is the way to do so.
.. NOTE::
If no local parameterizations are used, then the size of
the gradient vector is the sum of the sizes of all the parameter
blocks. If a parameter block has a local parameterization, then
it contributes "LocalSize" entries to the gradient vector.
.. NOTE::
This function cannot be called while the problem is being
solved, for example it cannot be called from an
:class:`IterationCallback` at the end of an iteration during a
solve.
.. class:: Problem::EvaluateOptions
Options struct that is used to control :func:`Problem::Evaluate`.
.. member:: vector<double*> Problem::EvaluateOptions::parameter_blocks
The set of parameter blocks for which evaluation should be
performed. This vector determines the order in which parameter
blocks occur in the gradient vector and in the columns of the
jacobian matrix. If parameter_blocks is empty, then it is assumed
to be equal to a vector containing ALL the parameter
blocks. Generally speaking the ordering of the parameter blocks in
this case depends on the order in which they were added to the
problem and whether or not the user removed any parameter blocks.
**NOTE** This vector should contain the same pointers as the ones
used to add parameter blocks to the Problem. These parameter block
should NOT point to new memory locations. Bad things will happen if
you do.
.. member:: vector<ResidualBlockId> Problem::EvaluateOptions::residual_blocks
The set of residual blocks for which evaluation should be
performed. This vector determines the order in which the residuals
occur, and how the rows of the jacobian are ordered. If
residual_blocks is empty, then it is assumed to be equal to the
vector containing all the residual blocks.
.. member:: bool Problem::EvaluateOptions::apply_loss_function
Even though the residual blocks in the problem may contain loss
functions, setting apply_loss_function to false will turn off the
application of the loss function to the output of the cost
function. This is of use for example if the user wishes to analyse
the solution quality by studying the distribution of residuals
before and after the solve.
.. member:: int Problem::EvaluateOptions::num_threads
Number of threads to use. (Requires OpenMP).
``rotation.h``
==============
Many applications of Ceres Solver involve optimization problems where
some of the variables correspond to rotations. To ease the pain of
work with the various representations of rotations (angle-axis,
quaternion and matrix) we provide a handy set of templated
functions. These functions are templated so that the user can use them
within Ceres Solver's automatic differentiation framework.
.. function:: template <typename T> void AngleAxisToQuaternion(T const* angle_axis, T* quaternion)
Convert a value in combined axis-angle representation to a
quaternion.
The value ``angle_axis`` is a triple whose norm is an angle in radians,
and whose direction is aligned with the axis of rotation, and
``quaternion`` is a 4-tuple that will contain the resulting quaternion.
.. function:: template <typename T> void QuaternionToAngleAxis(T const* quaternion, T* angle_axis)
Convert a quaternion to the equivalent combined axis-angle
representation.
The value ``quaternion`` must be a unit quaternion - it is not
normalized first, and ``angle_axis`` will be filled with a value
whose norm is the angle of rotation in radians, and whose direction
is the axis of rotation.
.. function:: template <typename T, int row_stride, int col_stride> void RotationMatrixToAngleAxis(const MatrixAdapter<const T, row_stride, col_stride>& R, T * angle_axis)
.. function:: template <typename T, int row_stride, int col_stride> void AngleAxisToRotationMatrix(T const * angle_axis, const MatrixAdapter<T, row_stride, col_stride>& R)
.. function:: template <typename T> void RotationMatrixToAngleAxis(T const * R, T * angle_axis)
.. function:: template <typename T> void AngleAxisToRotationMatrix(T const * angle_axis, T * R)
Conversions between 3x3 rotation matrix with given column and row strides and
axis-angle rotation representations. The functions that take a pointer to T instead
of a MatrixAdapter assume a column major representation with unit row stride and a column stride of 3.
.. function:: template <typename T, int row_stride, int col_stride> void EulerAnglesToRotationMatrix(const T* euler, const MatrixAdapter<T, row_stride, col_stride>& R)
.. function:: template <typename T> void EulerAnglesToRotationMatrix(const T* euler, int row_stride, T* R)
Conversions between 3x3 rotation matrix with given column and row strides and
Euler angle (in degrees) rotation representations.
The {pitch,roll,yaw} Euler angles are rotations around the {x,y,z}
axes, respectively. They are applied in that same order, so the
total rotation R is Rz * Ry * Rx.
The function that takes a pointer to T as the rotation matrix assumes a row
major representation with unit column stride and a row stride of 3.
The additional parameter row_stride is required to be 3.
.. function:: template <typename T, int row_stride, int col_stride> void QuaternionToScaledRotation(const T q[4], const MatrixAdapter<T, row_stride, col_stride>& R)
.. function:: template <typename T> void QuaternionToScaledRotation(const T q[4], T R[3 * 3])
Convert a 4-vector to a 3x3 scaled rotation matrix.
The choice of rotation is such that the quaternion
:math:`\begin{bmatrix} 1 &0 &0 &0\end{bmatrix}` goes to an identity
matrix and for small :math:`a, b, c` the quaternion
:math:`\begin{bmatrix}1 &a &b &c\end{bmatrix}` goes to the matrix
.. math::
I + 2 \begin{bmatrix} 0 & -c & b \\ c & 0 & -a\\ -b & a & 0
\end{bmatrix} + O(q^2)
which corresponds to a Rodrigues approximation, the last matrix
being the cross-product matrix of :math:`\begin{bmatrix} a& b&
c\end{bmatrix}`. Together with the property that :math:`R(q1 * q2)
= R(q1) * R(q2)` this uniquely defines the mapping from :math:`q` to
:math:`R`.
In the function that accepts a pointer to T instead of a MatrixAdapter,
the rotation matrix ``R`` is a row-major matrix with unit column stride
and a row stride of 3.
No normalization of the quaternion is performed, i.e.
:math:`R = \|q\|^2 Q`, where :math:`Q` is an orthonormal matrix
such that :math:`\det(Q) = 1` and :math:`Q*Q' = I`.
.. function:: template <typename T> void QuaternionToRotation(const T q[4], const MatrixAdapter<T, row_stride, col_stride>& R)
.. function:: template <typename T> void QuaternionToRotation(const T q[4], T R[3 * 3])
Same as above except that the rotation matrix is normalized by the
Frobenius norm, so that :math:`R R' = I` (and :math:`\det(R) = 1`).
.. function:: template <typename T> void UnitQuaternionRotatePoint(const T q[4], const T pt[3], T result[3])
Rotates a point pt by a quaternion q:
.. math:: \text{result} = R(q) \text{pt}
Assumes the quaternion is unit norm. If you pass in a quaternion
with :math:`|q|^2 = 2` then you WILL NOT get back 2 times the
result you get for a unit quaternion.
.. function:: template <typename T> void QuaternionRotatePoint(const T q[4], const T pt[3], T result[3])
With this function you do not need to assume that :math:`q` has unit norm.
It does assume that the norm is non-zero.
.. function:: template <typename T> void QuaternionProduct(const T z[4], const T w[4], T zw[4])
.. math:: zw = z * w
where :math:`*` is the Quaternion product between 4-vectors.
.. function:: template <typename T> void CrossProduct(const T x[3], const T y[3], T x_cross_y[3])
.. math:: \text{x_cross_y} = x \times y
.. function:: template <typename T> void AngleAxisRotatePoint(const T angle_axis[3], const T pt[3], T result[3])
.. math:: y = R(\text{angle_axis}) x
Cubic Interpolation
===================
Optimization problems often involve functions that are given in the
form of a table of values, for example an image. Evaluating these
functions and their derivatives requires interpolating these
values. Interpolating tabulated functions is a vast area of research
and there are a lot of libraries which implement a variety of
interpolation schemes. However, using them within the automatic
differentiation framework in Ceres is quite painful. To this end,
Ceres provides the ability to interpolate one dimensional and two
dimensional tabular functions.
The one dimensional interpolation is based on the Cubic Hermite
Spline, also known as the Catmull-Rom Spline. This produces a first
order differentiable interpolating function. The two dimensional
interpolation scheme is a generalization of the one dimensional scheme
where the interpolating function is assumed to be separable in the two
dimensions,
More details of the construction can be found `Linear Methods for
Image Interpolation <http://www.ipol.im/pub/art/2011/g_lmii/>`_ by
Pascal Getreuer.
.. class:: CubicInterpolator
Given as input an infinite one dimensional grid, which provides the
following interface.
.. code::
struct Grid1D {
enum { DATA_DIMENSION = 2; };
void GetValue(int n, double* f) const;
};
Where, ``GetValue`` gives us the value of a function :math:`f`
(possibly vector valued) for any integer :math:`n` and the enum
``DATA_DIMENSION`` indicates the dimensionality of the function being
interpolated. For example if you are interpolating rotations in
axis-angle format over time, then ``DATA_DIMENSION = 3``.
:class:`CubicInterpolator` uses Cubic Hermite splines to produce a
smooth approximation to it that can be used to evaluate the
:math:`f(x)` and :math:`f'(x)` at any point on the real number
line. For example, the following code interpolates an array of four
numbers.
.. code::
const double data[] = {1.0, 2.0, 5.0, 6.0};
Grid1D<double, 1> array(x, 0, 4);
CubicInterpolator interpolator(array);
double f, dfdx;
interpolator.Evaluate(1.5, &f, &dfdx);
In the above code we use ``Grid1D`` a templated helper class that
allows easy interfacing between ``C++`` arrays and
:class:`CubicInterpolator`.
``Grid1D`` supports vector valued functions where the various
coordinates of the function can be interleaved or stacked. It also
allows the use of any numeric type as input, as long as it can be
safely cast to a double.
.. class:: BiCubicInterpolator
Given as input an infinite two dimensional grid, which provides the
following interface:
.. code::
struct Grid2D {
enum { DATA_DIMENSION = 2 };
void GetValue(int row, int col, double* f) const;
};
Where, ``GetValue`` gives us the value of a function :math:`f`
(possibly vector valued) for any pair of integers :code:`row` and
:code:`col` and the enum ``DATA_DIMENSION`` indicates the
dimensionality of the function being interpolated. For example if you
are interpolating a color image with three channels (Red, Green &
Blue), then ``DATA_DIMENSION = 3``.
:class:`BiCubicInterpolator` uses the cubic convolution interpolation
algorithm of R. Keys [Keys]_, to produce a smooth approximation to it
that can be used to evaluate the :math:`f(r,c)`, :math:`\frac{\partial
f(r,c)}{\partial r}` and :math:`\frac{\partial f(r,c)}{\partial c}` at
any any point in the real plane.
For example the following code interpolates a two dimensional array.
.. code::
const double data[] = {1.0, 3.0, -1.0, 4.0,
3.6, 2.1, 4.2, 2.0,
2.0, 1.0, 3.1, 5.2};
Grid2D<double, 1> array(data, 0, 3, 0, 4);
BiCubicInterpolator interpolator(array);
double f, dfdr, dfdc;
interpolator.Evaluate(1.2, 2.5, &f, &dfdr, &dfdc);
In the above code, the templated helper class ``Grid2D`` is used to
make a ``C++`` array look like a two dimensional table to
:class:`BiCubicInterpolator`.
``Grid2D`` supports row or column major layouts. It also supports
vector valued functions where the individual coordinates of the
function may be interleaved or stacked. It also allows the use of any
numeric type as input, as long as it can be safely cast to double.