Feature extraction

opus_sources.mk

 OPUS_SOURCES = src/opus_decoder.c \
-src/opus_encoder.c
+src/opus_encoder.c \
+src/features.c

src/features.c

@@ -32,19 +32,44 @@
 #include "kiss_fft.h"
 #include "celt.h"
 #include "modes.h"
+#include "arch.h"
+#include "features.h"
+#include "quant_bands.h"
 
 #define NBANDS 17
 const int bands[NBANDS+1] =
 {1,  4,  8, 12, 16, 20, 24, 28, 32, 40, 48, 56, 64, 80, 96,112,136,160};
 
-void feature_analysis(CELTEncoder *celt_enc, celt_word16 *x)
+float dct_table[128] = {
+        0.250000, 0.250000, 0.250000, 0.250000, 0.250000, 0.250000, 0.250000, 0.250000,
+        0.250000, 0.250000, 0.250000, 0.250000, 0.250000, 0.250000, 0.250000, 0.250000,
+        0.351851, 0.338330, 0.311806, 0.273300, 0.224292, 0.166664, 0.102631, 0.034654,
+        -0.034654, -0.102631, -0.166664, -0.224292, -0.273300, -0.311806, -0.338330, -0.351851,
+        0.346760, 0.293969, 0.196424, 0.068975, -0.068975, -0.196424, -0.293969, -0.346760,
+        -0.346760, -0.293969, -0.196424, -0.068975, 0.068975, 0.196424, 0.293969, 0.346760,
+        0.338330, 0.224292, 0.034654, -0.166664, -0.311806, -0.351851, -0.273300, -0.102631,
+        0.102631, 0.273300, 0.351851, 0.311806, 0.166664, -0.034654, -0.224292, -0.338330,
+        0.326641, 0.135299, -0.135299, -0.326641, -0.326641, -0.135299, 0.135299, 0.326641,
+        0.326641, 0.135299, -0.135299, -0.326641, -0.326641, -0.135299, 0.135299, 0.326641,
+        0.311806, 0.034654, -0.273300, -0.338330, -0.102631, 0.224292, 0.351851, 0.166664,
+        -0.166664, -0.351851, -0.224292, 0.102631, 0.338330, 0.273300, -0.034654, -0.311806,
+        0.293969, -0.068975, -0.346760, -0.196424, 0.196424, 0.346760, 0.068975, -0.293969,
+        -0.293969, 0.068975, 0.346760, 0.196424, -0.196424, -0.346760, -0.068975, 0.293969,
+        0.273300, -0.166664, -0.338330, 0.034654, 0.351851, 0.102631, -0.311806, -0.224292,
+        0.224292, 0.311806, -0.102631, -0.351851, -0.034654, 0.338330, 0.166664, -0.273300,
+};
+
+static void feature_analysis(CELTEncoder *celt_enc, const celt_word16 *x,
+        celt_word16 *features, celt_word16 *mem)
 {
     int i;
     const CELTMode *mode;
     const kiss_fft_state *kfft;
     kiss_fft_cpx in[480], out[480];
     const celt_word16 *window;
-    celt_word32 E[NBANDS+1];
+    celt_word32 E[NBANDS];
+    celt_word16 logE[NBANDS];
+    celt_word16 BFCC[16];
     int overlap = 120;
     int N = 480;
 
@@ -70,11 +95,62 @@ void feature_analysis(CELTEncoder *celt_enc, celt_word16 *x)
     for (i=0;i<NBANDS;i++)
     {
         int j;
-        E[i] = 0;
+        celt_word32 sum = 0;
         for (j=bands[i];j<bands[i+1];j++)
-            E[i] = E[i] + MULT_32_32_Q31(out[  j].r, out[  j].r)
-                        + MULT_32_32_Q31(out[  j].i, out[  j].i)
-                        + MULT_32_32_Q31(out[N-j].r, out[N-j].r)
-                        + MULT_32_32_Q31(out[N-j].i, out[N-j].i);
+            sum = sum + MULT32_32_Q31(out[  j].r, out[  j].r)
+                      + MULT32_32_Q31(out[  j].i, out[  j].i)
+                      + MULT32_32_Q31(out[N-j].r, out[N-j].r)
+                      + MULT32_32_Q31(out[N-j].i, out[N-j].i);
+        E[i] = MAX32(EPSILON, sum);
+        //printf ("%f ", E[i]);
+    }
+    amp2Log2(mode, NBANDS, NBANDS, E, logE, 1);
+    for (i=0;i<NBANDS;i++)
+        logE[i] = MAX32(logE[i], -14.);
+    //for (i=0;i<16;i++)
+    //    printf ("%f ", logE[i]);
+
+    for (i=0;i<8;i++)
+    {
+        int j;
+        float sum = 0;
+        for (j=0;j<16;j++)
+            sum += dct_table[i*16+j]*logE[j];
+        BFCC[i] = sum;
+        //printf ("%f ", BFCC[i]);
+    }
+    for (i=0;i<7;i++)
+        features[i] = BFCC[i+1];
+
+    for (i=0;i<8;i++)
+        features[7+i] = .707*(BFCC[i] - mem[i+8]);
+    for (i=0;i<8;i++)
+        features[15+i] = .5*(BFCC[i] - 2*mem[i+8] + mem[i]);
+    for (i=0;i<8;i++)
+    {
+        mem[i+8] = mem[i];
+        mem[i] = BFCC[i];
     }
+    for (i=0;i<23;i++)
+        printf ("%f ", features[i]);
+
+    printf("\n");
+}
+
+void feature_analysis_fixed(CELTEncoder *celt_enc, const celt_int16 *x)
+{
+    /* FIXME: Get rid of this static var ASAP! */
+    static float mem[16];
+    float features[23];
+#ifdef FIXED_POINT
+    feature_analysis(celt_enc, x);
+#else
+    int i;
+    int N = 960-120;
+    celt_word16 x2[960-120];
+
+    for (i=0;i<N;i++)
+        x2[i] = x[i];
+    feature_analysis(celt_enc, x2, features, mem);
+#endif
 }

src/opus_encoder.c

@@ -38,6 +38,7 @@
 #include "entenc.h"
 #include "modes.h"
 #include "silk_API.h"
+#include "features.h"
 
 /* Transition tables for the voice and audio modes. First column is the
    middle (memoriless) threshold. The second column is the hysteresis
@@ -180,6 +181,8 @@ int opus_encode(OpusEncoder *st, const short *pcm, int frame_size,
     silk_enc = (char*)st+st->silk_enc_offset;
     celt_enc = (CELTEncoder*)((char*)st+st->celt_enc_offset);
 
+    feature_analysis_fixed(celt_enc, pcm);
+
     if (st->user_bitrate_bps==OPUS_BITRATE_AUTO)
         st->bitrate_bps = 60*st->Fs/frame_size + st->Fs*st->channels;
     else