-
Notifications
You must be signed in to change notification settings - Fork 6
/
audio.cpp
311 lines (257 loc) · 10.2 KB
/
audio.cpp
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
/* Edge Impulse Linux SDK
* Copyright (c) 2021 EdgeImpulse Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to deal
* in the Software without restriction, including without limitation the rights
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
* copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
*/
#include <stdlib.h>
#include <stdint.h>
#include <stdio.h>
#include <pthread.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <unistd.h>
#include <signal.h>
#include "edge-impulse-sdk/classifier/ei_run_classifier.h"
#include <alsa/asoundlib.h>
// Forward declarations
int microphone_audio_signal_get_data(size_t, size_t, float *);
static bool use_debug = false; // Set this to true to see e.g. features generated from the raw signal and log WAV files
static int16_t classifier_buffer[EI_CLASSIFIER_RAW_SAMPLE_COUNT * sizeof(int16_t)]; // full classifier buffer (used for debug)
static int16_t classifier_slice_buffer[EI_CLASSIFIER_SLICE_SIZE * sizeof(int16_t)]; // slice classifier buffer (used to classify)
// libalsa state
static snd_pcm_t *capture_handle;
static int channels = 1;
static unsigned int rate = EI_CLASSIFIER_FREQUENCY;
static snd_pcm_format_t format = SND_PCM_FORMAT_S16_LE;
static char *card;
/**
* Initialize the alsa library
*/
int init_alsa(bool debug = false) {
int err;
snd_pcm_hw_params_t *hw_params;
if ((err = snd_pcm_open(&capture_handle, card, SND_PCM_STREAM_CAPTURE, 0)) < 0) {
fprintf(stderr, "cannot open audio device %s (%s)\n",
card,
snd_strerror(err));
return 1;
}
if (debug) {
fprintf(stdout, "audio interface opened\n");
}
if ((err = snd_pcm_hw_params_malloc(&hw_params)) < 0) {
fprintf(stderr, "cannot allocate hardware parameter structure (%s)\n",
snd_strerror(err));
return 1;
}
if (debug) {
fprintf(stdout, "hw_params allocated\n");
}
if ((err = snd_pcm_hw_params_any(capture_handle, hw_params)) < 0)
{
fprintf(stderr, "cannot initialize hardware parameter structure (%s)\n",
snd_strerror(err));
return 1;
}
if (debug) {
fprintf(stdout, "hw_params initialized\n");
}
if ((err = snd_pcm_hw_params_set_access(capture_handle, hw_params, SND_PCM_ACCESS_RW_INTERLEAVED)) < 0)
{
fprintf(stderr, "cannot set access type (%s)\n",
snd_strerror(err));
return 1;
}
if (debug) {
fprintf(stdout, "hw_params access set\n");
}
if ((err = snd_pcm_hw_params_set_format(capture_handle, hw_params, format)) < 0)
{
fprintf(stderr, "cannot set format (%s)\n",
snd_strerror(err));
return 1;
}
if (debug) {
fprintf(stdout, "hw_params format set\n");
}
if ((err = snd_pcm_hw_params_set_rate(capture_handle, hw_params, rate, 0)) < 0) {
fprintf(stderr, "cannot set sample rate (%s)\n",
snd_strerror(err));
return 1;
}
else {
unsigned int read_rate;
int read_dir;
snd_pcm_hw_params_get_rate(hw_params, &read_rate, &read_dir);
if (debug) {
fprintf(stdout, "hw_params rate set: %d\n", read_rate);
}
}
if ((err = snd_pcm_hw_params_set_channels(capture_handle, hw_params, channels)) < 0) {
fprintf(stderr, "cannot set channel count (%s)\n",
snd_strerror(err));
return 1;
}
if (debug) {
fprintf(stdout, "hw_params channels set:%d\n", channels);
}
if ((err = snd_pcm_hw_params(capture_handle, hw_params)) < 0) {
fprintf(stderr, "cannot set parameters (%s)\n",
snd_strerror(err));
return 1;
}
if (debug) {
fprintf(stdout, "hw_params set\n");
}
snd_pcm_hw_params_free(hw_params);
if (debug) {
fprintf(stdout, "hw_params freed\n");
}
if ((err = snd_pcm_prepare(capture_handle)) < 0)
{
fprintf(stderr, "cannot prepare audio interface for use (%s)\n",
snd_strerror(err));
return 1;
}
if (debug) {
fprintf(stdout, "audio interface prepared\n");
}
return 0;
}
void close_alsa(int signum) {
snd_pcm_drop(capture_handle);
snd_pcm_close(capture_handle);
exit(0);
}
/**
* Classify the current buffer
*/
void classify_current_buffer() {
// write the WAV file for debug purposes...
if (use_debug) {
char filename[128] = { 0 };
static int classify_counter = 0;
struct stat st = { 0 };
if (stat("out", &st) == -1) {
mkdir("out", 0700);
}
uint32_t wavFreq = EI_CLASSIFIER_FREQUENCY;
uint32_t fileSize = 44 + (EI_CLASSIFIER_RAW_SAMPLE_COUNT * sizeof(int16_t));
uint32_t dataSize = (EI_CLASSIFIER_RAW_SAMPLE_COUNT * sizeof(int16_t));
uint32_t srBpsC8 = (wavFreq * 16 * 1) / 8;
uint8_t wav_header[44] = {
0x52, 0x49, 0x46, 0x46, // RIFF
(uint8_t)(fileSize & 0xff), (uint8_t)((fileSize >> 8) & 0xff), (uint8_t)((fileSize >> 16) & 0xff), (uint8_t)((fileSize >> 24) & 0xff),
0x57, 0x41, 0x56, 0x45, // WAVE
0x66, 0x6d, 0x74, 0x20, // fmt
0x10, 0x00, 0x00, 0x00, // length of format data
0x01, 0x00, // type of format (1=PCM)
0x01, 0x00, // number of channels
(uint8_t)(wavFreq & 0xff), (uint8_t)((wavFreq >> 8) & 0xff), (uint8_t)((wavFreq >> 16) & 0xff), (uint8_t)((wavFreq >> 24) & 0xff),
(uint8_t)(srBpsC8 & 0xff), (uint8_t)((srBpsC8 >> 8) & 0xff), (uint8_t)((srBpsC8 >> 16) & 0xff), (uint8_t)((srBpsC8 >> 24) & 0xff),
0x02, 0x00, 0x10, 0x00,
0x64, 0x61, 0x74, 0x61, // data
(uint8_t)(dataSize & 0xff), (uint8_t)((dataSize >> 8) & 0xff), (uint8_t)((dataSize >> 16) & 0xff), (uint8_t)((dataSize >> 24) & 0xff),
};
snprintf(filename, 128, "out/data.%d.wav", ++classify_counter);
FILE *f = fopen(filename, "w+");
if (!f) {
printf("Failed to create file '%s'\n", filename);
return;
}
fwrite(wav_header, 1, 44, f);
fwrite(classifier_buffer, 2, EI_CLASSIFIER_RAW_SAMPLE_COUNT, f);
fclose(f);
}
// classify the current buffer and print the results
signal_t signal;
signal.total_length = EI_CLASSIFIER_SLICE_SIZE;
signal.get_data = µphone_audio_signal_get_data;
ei_impulse_result_t result = { 0 };
EI_IMPULSE_ERROR r = run_classifier_continuous(&signal, &result, use_debug, true);
if (r != EI_IMPULSE_OK) {
printf("ERR: Failed to run classifier (%d)\n", r);
return;
}
printf("%d ms. ", result.timing.dsp + result.timing.classification);
for (size_t ix = 0; ix < EI_CLASSIFIER_LABEL_COUNT; ix++) {
printf("%s: %.05f", result.classification[ix].label, result.classification[ix].value);
if (ix != EI_CLASSIFIER_LABEL_COUNT - 1) {
printf(", ");
}
}
printf("\n");
}
/**
* @brief main function. Runs the inferencing loop.
*/
int main(int argc, char **argv)
{
if (argc < 2) {
printf("Requires one parameter (ID of the sound card) in the form of plughw:1,0 (where 1=card number, 0=device).\n");
printf("You can find these via `cat /proc/asound/cards`. E.g. for:\n");
printf(" 0 [Headphones ]: bcm2835_headphonbcm2835 Headphones - bcm2835 Headphones\n");
printf(" bcm2835 Headphones\n");
printf(" 1 [Webcam ]: USB-Audio - C922 Pro Stream Webcam\n");
printf(" C922 Pro Stream Webcam at usb-0000:01:00.0-1.3, high speed\n");
printf("The ID for 'C922 Pro Stream Webcam' is then plughw:1,0\n");
exit(1);
}
card = argv[1];
for (int ix = 2; ix < argc; ix++) {
if (strcmp(argv[ix], "--debug") == 0) {
printf("Enabling debug mode\n");
use_debug = true;
}
}
if (init_alsa(use_debug) != 0) {
exit(1);
}
::signal(SIGINT, close_alsa);
run_classifier_init();
while (1) {
int x = snd_pcm_readi(capture_handle, classifier_slice_buffer, EI_CLASSIFIER_SLICE_SIZE);
if (x != EI_CLASSIFIER_SLICE_SIZE) {
printf("Failed to read audio data (%d)\n", x);
return 1;
}
// so let's say we have a 16000 element classifier_buffer
// then we want to move 4000..16000 to position 0..12000
// and fill 12000..16000 with the data from classifier_slice_buffer
// (this is only used to write full second length audio files in debug mode,
// the actual inference is done just on the slice)
// 1. roll -EI_CLASSIFIER_SLICE_SIZE here
numpy::roll(classifier_buffer, EI_CLASSIFIER_RAW_SAMPLE_COUNT, -EI_CLASSIFIER_SLICE_SIZE);
// 2. copy slice buffer to the end
const size_t classifier_buffer_offset = EI_CLASSIFIER_RAW_SAMPLE_COUNT - EI_CLASSIFIER_SLICE_SIZE;
memcpy(classifier_buffer + classifier_buffer_offset, classifier_slice_buffer, EI_CLASSIFIER_SLICE_SIZE * sizeof(int16_t));
// 3. and classify!
classify_current_buffer();
}
close_alsa(0);
}
/**
* Get data from the classifier buffer
*/
int microphone_audio_signal_get_data(size_t offset, size_t length, float *out_ptr) {
return numpy::int16_to_float(classifier_slice_buffer + offset, out_ptr, length);
}
#if !defined(EI_CLASSIFIER_SENSOR) || EI_CLASSIFIER_SENSOR != EI_CLASSIFIER_SENSOR_MICROPHONE
#error "Invalid model for current sensor."
#endif