Line data Source code
1 : /* Copyright (c) 2014, Cisco Systems, INC
2 : Written by XiangMingZhu WeiZhou MinPeng YanWang
3 :
4 : Redistribution and use in source and binary forms, with or without
5 : modification, are permitted provided that the following conditions
6 : are met:
7 :
8 : - Redistributions of source code must retain the above copyright
9 : notice, this list of conditions and the following disclaimer.
10 :
11 : - Redistributions in binary form must reproduce the above copyright
12 : notice, this list of conditions and the following disclaimer in the
13 : documentation and/or other materials provided with the distribution.
14 :
15 : THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
16 : ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
17 : LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
18 : A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER
19 : OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
20 : EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
21 : PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
22 : PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
23 : LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
24 : NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
25 : SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
26 : */
27 :
28 : #ifdef HAVE_CONFIG_H
29 : #include "config.h"
30 : #endif
31 :
32 : #include <xmmintrin.h>
33 : #include <emmintrin.h>
34 : #include <smmintrin.h>
35 :
36 : #include "main.h"
37 : #include "stack_alloc.h"
38 :
39 : /* Weighting factors for tilt measure */
40 : static const opus_int32 tiltWeights[ VAD_N_BANDS ] = { 30000, 6000, -12000, -12000 };
41 :
42 : /***************************************/
43 : /* Get the speech activity level in Q8 */
44 : /***************************************/
45 0 : opus_int silk_VAD_GetSA_Q8_sse4_1( /* O Return value, 0 if success */
46 : silk_encoder_state *psEncC, /* I/O Encoder state */
47 : const opus_int16 pIn[] /* I PCM input */
48 : )
49 : {
50 : opus_int SA_Q15, pSNR_dB_Q7, input_tilt;
51 : opus_int decimated_framelength1, decimated_framelength2;
52 : opus_int decimated_framelength;
53 : opus_int dec_subframe_length, dec_subframe_offset, SNR_Q7, i, b, s;
54 : opus_int32 sumSquared, smooth_coef_Q16;
55 : opus_int16 HPstateTmp;
56 : VARDECL( opus_int16, X );
57 : opus_int32 Xnrg[ VAD_N_BANDS ];
58 : opus_int32 NrgToNoiseRatio_Q8[ VAD_N_BANDS ];
59 : opus_int32 speech_nrg, x_tmp;
60 : opus_int X_offset[ VAD_N_BANDS ];
61 0 : opus_int ret = 0;
62 0 : silk_VAD_state *psSilk_VAD = &psEncC->sVAD;
63 :
64 : SAVE_STACK;
65 :
66 : /* Safety checks */
67 : silk_assert( VAD_N_BANDS == 4 );
68 0 : silk_assert( MAX_FRAME_LENGTH >= psEncC->frame_length );
69 0 : silk_assert( psEncC->frame_length <= 512 );
70 0 : silk_assert( psEncC->frame_length == 8 * silk_RSHIFT( psEncC->frame_length, 3 ) );
71 :
72 : /***********************/
73 : /* Filter and Decimate */
74 : /***********************/
75 0 : decimated_framelength1 = silk_RSHIFT( psEncC->frame_length, 1 );
76 0 : decimated_framelength2 = silk_RSHIFT( psEncC->frame_length, 2 );
77 0 : decimated_framelength = silk_RSHIFT( psEncC->frame_length, 3 );
78 : /* Decimate into 4 bands:
79 : 0 L 3L L 3L 5L
80 : - -- - -- --
81 : 8 8 2 4 4
82 :
83 : [0-1 kHz| temp. |1-2 kHz| 2-4 kHz | 4-8 kHz |
84 :
85 : They're arranged to allow the minimal ( frame_length / 4 ) extra
86 : scratch space during the downsampling process */
87 0 : X_offset[ 0 ] = 0;
88 0 : X_offset[ 1 ] = decimated_framelength + decimated_framelength2;
89 0 : X_offset[ 2 ] = X_offset[ 1 ] + decimated_framelength;
90 0 : X_offset[ 3 ] = X_offset[ 2 ] + decimated_framelength2;
91 0 : ALLOC( X, X_offset[ 3 ] + decimated_framelength1, opus_int16 );
92 :
93 : /* 0-8 kHz to 0-4 kHz and 4-8 kHz */
94 0 : silk_ana_filt_bank_1( pIn, &psSilk_VAD->AnaState[ 0 ],
95 0 : X, &X[ X_offset[ 3 ] ], psEncC->frame_length );
96 :
97 : /* 0-4 kHz to 0-2 kHz and 2-4 kHz */
98 0 : silk_ana_filt_bank_1( X, &psSilk_VAD->AnaState1[ 0 ],
99 0 : X, &X[ X_offset[ 2 ] ], decimated_framelength1 );
100 :
101 : /* 0-2 kHz to 0-1 kHz and 1-2 kHz */
102 0 : silk_ana_filt_bank_1( X, &psSilk_VAD->AnaState2[ 0 ],
103 0 : X, &X[ X_offset[ 1 ] ], decimated_framelength2 );
104 :
105 : /*********************************************/
106 : /* HP filter on lowest band (differentiator) */
107 : /*********************************************/
108 0 : X[ decimated_framelength - 1 ] = silk_RSHIFT( X[ decimated_framelength - 1 ], 1 );
109 0 : HPstateTmp = X[ decimated_framelength - 1 ];
110 0 : for( i = decimated_framelength - 1; i > 0; i-- ) {
111 0 : X[ i - 1 ] = silk_RSHIFT( X[ i - 1 ], 1 );
112 0 : X[ i ] -= X[ i - 1 ];
113 : }
114 0 : X[ 0 ] -= psSilk_VAD->HPstate;
115 0 : psSilk_VAD->HPstate = HPstateTmp;
116 :
117 : /*************************************/
118 : /* Calculate the energy in each band */
119 : /*************************************/
120 0 : for( b = 0; b < VAD_N_BANDS; b++ ) {
121 : /* Find the decimated framelength in the non-uniformly divided bands */
122 0 : decimated_framelength = silk_RSHIFT( psEncC->frame_length, silk_min_int( VAD_N_BANDS - b, VAD_N_BANDS - 1 ) );
123 :
124 : /* Split length into subframe lengths */
125 0 : dec_subframe_length = silk_RSHIFT( decimated_framelength, VAD_INTERNAL_SUBFRAMES_LOG2 );
126 0 : dec_subframe_offset = 0;
127 :
128 : /* Compute energy per sub-frame */
129 : /* initialize with summed energy of last subframe */
130 0 : Xnrg[ b ] = psSilk_VAD->XnrgSubfr[ b ];
131 0 : for( s = 0; s < VAD_INTERNAL_SUBFRAMES; s++ ) {
132 : __m128i xmm_X, xmm_acc;
133 0 : sumSquared = 0;
134 :
135 0 : xmm_acc = _mm_setzero_si128();
136 :
137 0 : for( i = 0; i < dec_subframe_length - 7; i += 8 )
138 : {
139 0 : xmm_X = _mm_loadu_si128( (__m128i *)&(X[ X_offset[ b ] + i + dec_subframe_offset ] ) );
140 0 : xmm_X = _mm_srai_epi16( xmm_X, 3 );
141 0 : xmm_X = _mm_madd_epi16( xmm_X, xmm_X );
142 0 : xmm_acc = _mm_add_epi32( xmm_acc, xmm_X );
143 : }
144 :
145 0 : xmm_acc = _mm_add_epi32( xmm_acc, _mm_unpackhi_epi64( xmm_acc, xmm_acc ) );
146 0 : xmm_acc = _mm_add_epi32( xmm_acc, _mm_shufflelo_epi16( xmm_acc, 0x0E ) );
147 :
148 0 : sumSquared += _mm_cvtsi128_si32( xmm_acc );
149 :
150 0 : for( ; i < dec_subframe_length; i++ ) {
151 : /* The energy will be less than dec_subframe_length * ( silk_int16_MIN / 8 ) ^ 2. */
152 : /* Therefore we can accumulate with no risk of overflow (unless dec_subframe_length > 128) */
153 0 : x_tmp = silk_RSHIFT(
154 : X[ X_offset[ b ] + i + dec_subframe_offset ], 3 );
155 0 : sumSquared = silk_SMLABB( sumSquared, x_tmp, x_tmp );
156 :
157 : /* Safety check */
158 0 : silk_assert( sumSquared >= 0 );
159 : }
160 :
161 : /* Add/saturate summed energy of current subframe */
162 0 : if( s < VAD_INTERNAL_SUBFRAMES - 1 ) {
163 0 : Xnrg[ b ] = silk_ADD_POS_SAT32( Xnrg[ b ], sumSquared );
164 : } else {
165 : /* Look-ahead subframe */
166 0 : Xnrg[ b ] = silk_ADD_POS_SAT32( Xnrg[ b ], silk_RSHIFT( sumSquared, 1 ) );
167 : }
168 :
169 0 : dec_subframe_offset += dec_subframe_length;
170 : }
171 0 : psSilk_VAD->XnrgSubfr[ b ] = sumSquared;
172 : }
173 :
174 : /********************/
175 : /* Noise estimation */
176 : /********************/
177 0 : silk_VAD_GetNoiseLevels( &Xnrg[ 0 ], psSilk_VAD );
178 :
179 : /***********************************************/
180 : /* Signal-plus-noise to noise ratio estimation */
181 : /***********************************************/
182 0 : sumSquared = 0;
183 0 : input_tilt = 0;
184 0 : for( b = 0; b < VAD_N_BANDS; b++ ) {
185 0 : speech_nrg = Xnrg[ b ] - psSilk_VAD->NL[ b ];
186 0 : if( speech_nrg > 0 ) {
187 : /* Divide, with sufficient resolution */
188 0 : if( ( Xnrg[ b ] & 0xFF800000 ) == 0 ) {
189 0 : NrgToNoiseRatio_Q8[ b ] = silk_DIV32( silk_LSHIFT( Xnrg[ b ], 8 ), psSilk_VAD->NL[ b ] + 1 );
190 : } else {
191 0 : NrgToNoiseRatio_Q8[ b ] = silk_DIV32( Xnrg[ b ], silk_RSHIFT( psSilk_VAD->NL[ b ], 8 ) + 1 );
192 : }
193 :
194 : /* Convert to log domain */
195 0 : SNR_Q7 = silk_lin2log( NrgToNoiseRatio_Q8[ b ] ) - 8 * 128;
196 :
197 : /* Sum-of-squares */
198 0 : sumSquared = silk_SMLABB( sumSquared, SNR_Q7, SNR_Q7 ); /* Q14 */
199 :
200 : /* Tilt measure */
201 0 : if( speech_nrg < ( (opus_int32)1 << 20 ) ) {
202 : /* Scale down SNR value for small subband speech energies */
203 0 : SNR_Q7 = silk_SMULWB( silk_LSHIFT( silk_SQRT_APPROX( speech_nrg ), 6 ), SNR_Q7 );
204 : }
205 0 : input_tilt = silk_SMLAWB( input_tilt, tiltWeights[ b ], SNR_Q7 );
206 : } else {
207 0 : NrgToNoiseRatio_Q8[ b ] = 256;
208 : }
209 : }
210 :
211 : /* Mean-of-squares */
212 0 : sumSquared = silk_DIV32_16( sumSquared, VAD_N_BANDS ); /* Q14 */
213 :
214 : /* Root-mean-square approximation, scale to dBs, and write to output pointer */
215 0 : pSNR_dB_Q7 = (opus_int16)( 3 * silk_SQRT_APPROX( sumSquared ) ); /* Q7 */
216 :
217 : /*********************************/
218 : /* Speech Probability Estimation */
219 : /*********************************/
220 0 : SA_Q15 = silk_sigm_Q15( silk_SMULWB( VAD_SNR_FACTOR_Q16, pSNR_dB_Q7 ) - VAD_NEGATIVE_OFFSET_Q5 );
221 :
222 : /**************************/
223 : /* Frequency Tilt Measure */
224 : /**************************/
225 0 : psEncC->input_tilt_Q15 = silk_LSHIFT( silk_sigm_Q15( input_tilt ) - 16384, 1 );
226 :
227 : /**************************************************/
228 : /* Scale the sigmoid output based on power levels */
229 : /**************************************************/
230 0 : speech_nrg = 0;
231 0 : for( b = 0; b < VAD_N_BANDS; b++ ) {
232 : /* Accumulate signal-without-noise energies, higher frequency bands have more weight */
233 0 : speech_nrg += ( b + 1 ) * silk_RSHIFT( Xnrg[ b ] - psSilk_VAD->NL[ b ], 4 );
234 : }
235 :
236 : /* Power scaling */
237 0 : if( speech_nrg <= 0 ) {
238 0 : SA_Q15 = silk_RSHIFT( SA_Q15, 1 );
239 0 : } else if( speech_nrg < 32768 ) {
240 0 : if( psEncC->frame_length == 10 * psEncC->fs_kHz ) {
241 0 : speech_nrg = silk_LSHIFT_SAT32( speech_nrg, 16 );
242 : } else {
243 0 : speech_nrg = silk_LSHIFT_SAT32( speech_nrg, 15 );
244 : }
245 :
246 : /* square-root */
247 0 : speech_nrg = silk_SQRT_APPROX( speech_nrg );
248 0 : SA_Q15 = silk_SMULWB( 32768 + speech_nrg, SA_Q15 );
249 : }
250 :
251 : /* Copy the resulting speech activity in Q8 */
252 0 : psEncC->speech_activity_Q8 = silk_min_int( silk_RSHIFT( SA_Q15, 7 ), silk_uint8_MAX );
253 :
254 : /***********************************/
255 : /* Energy Level and SNR estimation */
256 : /***********************************/
257 : /* Smoothing coefficient */
258 0 : smooth_coef_Q16 = silk_SMULWB( VAD_SNR_SMOOTH_COEF_Q18, silk_SMULWB( (opus_int32)SA_Q15, SA_Q15 ) );
259 :
260 0 : if( psEncC->frame_length == 10 * psEncC->fs_kHz ) {
261 0 : smooth_coef_Q16 >>= 1;
262 : }
263 :
264 0 : for( b = 0; b < VAD_N_BANDS; b++ ) {
265 : /* compute smoothed energy-to-noise ratio per band */
266 0 : psSilk_VAD->NrgRatioSmth_Q8[ b ] = silk_SMLAWB( psSilk_VAD->NrgRatioSmth_Q8[ b ],
267 : NrgToNoiseRatio_Q8[ b ] - psSilk_VAD->NrgRatioSmth_Q8[ b ], smooth_coef_Q16 );
268 :
269 : /* signal to noise ratio in dB per band */
270 0 : SNR_Q7 = 3 * ( silk_lin2log( psSilk_VAD->NrgRatioSmth_Q8[b] ) - 8 * 128 );
271 : /* quality = sigmoid( 0.25 * ( SNR_dB - 16 ) ); */
272 0 : psEncC->input_quality_bands_Q15[ b ] = silk_sigm_Q15( silk_RSHIFT( SNR_Q7 - 16 * 128, 4 ) );
273 : }
274 :
275 : RESTORE_STACK;
276 0 : return( ret );
277 : }
|
