Edinburgh Speech Tools  2.1-release
 All Classes Files Functions Variables Typedefs Enumerations Enumerator Friends Macros Groups Pages
Functions
EST_Track Auxiliary functions

Functions

void track_smooth (EST_Track &c, float x, EST_String stype="")
 
void time_med_smooth (EST_Track &c, float x)
 
void time_mean_smooth (EST_Track &c, float x)
 
void simple_med_smooth (EST_Track &c, int n, int channel=0)
 
void simple_mean_smooth (EST_Track &c, int n, int channel=0)
 
float abs_error (EST_Track &a, EST_Track &b, int channel)
 
void absolute (EST_Track &tr)
 
void normalise (EST_Track &tr)
 
void normalise (EST_Track &tr, float mean, float sd, int channel, float upper, float lower)
 
void normalise (EST_Track &tr, EST_FVector &mean, EST_FVector &sd, float upper, float lower)
 
void normalise (EST_TrackList &trlist, EST_FVector &mean, EST_FVector &sd, float upper, float lower)
 
EST_Track differentiate (EST_Track &c, float samp_int=0.0)
 
EST_Track difference (EST_Track &a, EST_Track &b)
 
float mean (const EST_Track &a, int channel)
 
void mean (const EST_Track &a, EST_FVector &m)
 
void meansd (EST_Track &a, float &m, float &sd, int channel)
 
float rms_error (EST_Track &a, EST_Track &b, int channel)
 
float correlation (EST_Track &a, EST_Track &b, int channel)
 
void meansd (EST_Track &a, EST_FVector &m, EST_FVector &sd)
 
EST_FVector rms_error (EST_Track &a, EST_Track &b)
 
EST_FVector abs_error (EST_Track &a, EST_Track &b)
 
EST_FVector correlation (EST_Track &a, EST_Track &b)
 
void align_to_track (EST_Track &tr, float &start, float &end)
 Move the start and end variables to the nearest frame.
 
void align_to_track (EST_Track &tr, int &start, int &end, int sample_rate)
 Move the start and end variables to the nearest frame.
 
void move_to_frame_ends (EST_Track &tr, int &start, int &end, int sample_rate, float offset=0.0)
 Move the start and end variables to the start and end of the nearest frame.
 
int nearest_boundary (EST_Track &tr, float time, int sample_rate, float offset=0)
 Index of the frame whose start boundary.
 
void set_start (EST_Track &tr, float start)
 Move the track so that it starts at the indicated time.
 
void move_start (EST_Track &tr, float shift)
 Move the track by shift seconds.
 
EST_Track error (EST_Track &ref, EST_Track &test, int relax=0)
 
void extract (EST_Track &orig, float start, float end, EST_Track &res)
 
int track_divide (EST_TrackList &mtfr, EST_Track &fv, EST_Relation &key)
 
void ParallelTracks (EST_Track &a, EST_TrackList &list, const EST_String &style)
 
void track_info (EST_Track &track)
 
EST_String options_track_filetypes (void)
 
EST_String options_track_filetypes_long (void)
 
EST_String options_subtrack (void)
 
int read_track (EST_Track &tr, const EST_String &in_file, EST_Option &al)
 
float get_time_frame_size (EST_Track &pms, int i, int prefer_prev=0)
 
int get_frame_size (EST_Track &pms, int current_pos, int sample_rate, int prefer_prev=0)
 
int get_order (const EST_Track &t, EST_CoefficientType type, int d=0)
 How many coefficients in track (looks for Coef0 and coefN channels)
 
int get_order (const EST_Track &t)
 
int sum_lengths (const EST_Track &t, int sample_rate, int start_frame=0, int end_frame=-1)
 Total the length channel values.
 
void get_start_positions (const EST_Track &t, int sample_rate, EST_TBuffer< int > &pos)
 Find the start point in the signal of the sections of speech related to each frame.
 
void channel_to_time (EST_Track &tr, int channel, float scale=1.0)
 
void channel_to_time (EST_Track &tr, EST_ChannelType c, float scale=1.0)
 
void channel_to_time (EST_Track &tr, const EST_String c_name, float scale=1.0)
 
void channel_to_time_lengths (EST_Track &tr, int channel, float scale=1.0)
 
void channel_to_time_lengths (EST_Track &tr, EST_ChannelType c, float scale=1.0)
 
void channel_to_time_lengths (EST_Track &tr, const EST_String c_name, float scale=1.0)
 

Detailed Description

Author
Paul Taylor pault.nosp@m.@cst.nosp@m.r.ed..nosp@m.ac.u.nosp@m.k
Version
Id
EST_track_aux.h,v 1.4 2004/05/24 11:15:51 korin Exp

Function Documentation

float abs_error ( EST_Track a,
EST_Track b,
int  channel 
)

Calculate the mean absolute error between the same channel in two tracks. This is given by

\[\frac{1}{n}\sum_{i=1}^{n}|a_{i} - b_{i}|\]

Parameters
aFirst track
bSecond track
channelChannel from track for which the error is to be computed
Returns
absolute error
See also
abs_error, rms_error(EST_Track &a, EST_Track &b)

Definition at line 410 of file EST_track_aux.cc.

EST_Track differentiate ( EST_Track c,
float  samp_int = 0.0 
)

Calculate the simple derivative of a track. This is given by

\[a_{i+1} - a_{i}\]

The values in the resultant track are spaced midway between the values in the input track, resulting in 1 fewer frames in the track. This is a very local estimation of the derivative of the track at a point in time. A smoother value can be obtained using the delta function.

See also
delta

Definition at line 249 of file EST_track_aux.cc.

void meansd ( EST_Track tr,
float &  m,
float &  sd,
int  channel 
)

Calculate the mead and standard deviation for a single channel of a track

Definition at line 370 of file EST_track_aux.cc.

float rms_error ( EST_Track a,
EST_Track b,
int  channel 
)

Calculate the root mean square error between the same channel in two tracks. The channel is identified by its index.

See also
abs_error, float rms_error(EST_Track &a, EST_Track &b)

Calculate the root mean square error between the same channel in two tracks

See also
abs_error, rms_error

Definition at line 396 of file EST_track_aux.cc.

EST_FVector rms_error ( EST_Track a,
EST_Track b 
)

Calculate the root mean square error between each channels in two tracks. For two tracks of M channels, the result is returned as an EST_FVector of size M, with element i representing the rms error for channel i.

See also
abs_error, rms_error

Definition at line 513 of file EST_track_aux.cc.

float get_time_frame_size ( EST_Track pms,
int  i,
int  prefer_prev = 0 
)

Return the frame size in seconds based on analysis of current time points. This function basically determines the local frame size (shift) by subtracting the current time point from the next time point. If the prefer_prev flag is set to true, or the index is the last in the track, the size is determined by subtracting the previous time point from the current one.

This is most commonly used in pitch synchronous analysis to determine the local pitch period.

See also
get_frame_size

Definition at line 331 of file sigpr_utt.cc.

int get_frame_size ( EST_Track pms,
int  current_pos,
int  sample_rate,
int  prefer_prev = 0 
)

Return the frame size in samples based on analysis of current time points. This function basically determines the local frame size (shift) by subtracting the current time point from the next time point. If the prefer_prev flag is set to true, or the index is the last in the track, the size is determined by subtracting the previous time point from the current one.

This is most commonly used in pitch synchronous analysis to determine the local pitch period.

See also
get_time_frame_size

Definition at line 315 of file sigpr_utt.cc.