Public Member Functions |
| scalerclass_energy (bool apply_tanh=false, uint jsize=1, uint jitter_selection=0, float dist_coeff=1.0, float scale_coeff=1.0, bool predict_conf=false, bool predict_bconf=false, idx< T > *biases=NULL, idx< T > *coeffs=NULL, const char *name="scalerclass_energy") |
virtual | ~scalerclass_energy () |
| destructor.
|
virtual void | fprop (Tstate &in, Tstate &scale, Tstate &energy) |
| fprop: compute output from input
|
virtual void | bprop (Tstate &in, Tstate &scale, Tstate &energy) |
| bprop: compute gradient wrt inputs, given gradient wrt output
|
virtual void | bbprop (Tstate &in, Tstate &scale, Tstate &energy) |
| bprop: compute diaghession wrt inputs, given diaghessian wrt output
|
virtual void | infer2 (Tstate &i1, Tstate &scale, infer_param &ip, Tstate *energy=NULL) |
| compute value of in2 that minimizes the energy, given in1
|
virtual std::string | describe () |
| Returns a string describing this module and its parameters.
|
Public Attributes |
idx< T > | last_target_raw |
| Un-normalized target.
|
Tstate | last_target |
| Last target selected by fprop.
|
Protected Attributes |
uint | jsize |
| Number of jitter elements.
|
bool | apply_tanh |
| If true, apply tanh to inputs.
|
uint | jitter_selection |
| Selection type.
|
float | dist_coeff |
| Coefficient of distance score.
|
float | scale_coeff |
| Coefficient of scale score.
|
Tstate | tmp |
| Temporary buffer.
|
Tstate | tmp2 |
| Temporary buffer.
|
Tstate | last_class_target |
| Last class target selected.
|
Tstate | last_jitt_target |
| Last jitter target selected.
|
Tstate | last_conf_target |
| Last confidence target selected.
|
idx< T > | best_target |
| The last target selected by fprop.
|
tanh_module< T, Tstate > | mtanh |
| A tanh module.
|
bool | predict_conf |
| Predict confidence.
|
bool | predict_bconf |
| Predicted confidence is binary or not.
|
idx< T > * | biases |
| Normalization biases: scale,h,w,conf.
|
idx< T > * | coeffs |
| Normalization coeffs: scale,h,w,conf.
|