Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -188,8 +188,9 @@ def add_voice_labelv2(json_file, audio_path):
|
|
| 188 |
speech_prob = model(chunk, SAMPLING_RATE).item()
|
| 189 |
speech_probs.append(speech_prob)
|
| 190 |
vad_iterator.reset_states() # reset model states after each audio
|
| 191 |
-
|
| 192 |
voice_idxs = np.where(np.array(speech_probs) >= 0.7)
|
|
|
|
| 193 |
if len(voice_idxs) == 0:
|
| 194 |
print("NO VOICE SEGMENTS DETECTED!")
|
| 195 |
try:
|
|
@@ -202,7 +203,7 @@ def add_voice_labelv2(json_file, audio_path):
|
|
| 202 |
begin_seq = False
|
| 203 |
if voice_idxs[i+1] == voice_idxs[i]+1:
|
| 204 |
continue
|
| 205 |
-
|
| 206 |
start_time = float((start_idx*window_size_samples)/SAMPLING_RATE)
|
| 207 |
end_time = float((voice_idxs[i]*window_size_samples)/SAMPLING_RATE)
|
| 208 |
|
|
@@ -210,7 +211,8 @@ def add_voice_labelv2(json_file, audio_path):
|
|
| 210 |
end_minutes = int(end_time)
|
| 211 |
start_seconds = (start_time - start_minutes) * 60
|
| 212 |
end_seconds = (end_time - end_minutes) * 60
|
| 213 |
-
|
|
|
|
| 214 |
data['vocal_times'] = {
|
| 215 |
"start_time": f"{start_minutes}.{start_seconds:.2f}",
|
| 216 |
"end_time": f"{end_minutes}.{end_seconds:.2f}"
|
|
@@ -224,6 +226,7 @@ def add_voice_labelv2(json_file, audio_path):
|
|
| 224 |
print(f"An exception occurred: {e}")
|
| 225 |
|
| 226 |
with open(json_file, 'w') as f:
|
|
|
|
| 227 |
json.dump(data, f, indent=4)
|
| 228 |
|
| 229 |
with gr.Blocks() as demo:
|
|
|
|
| 188 |
speech_prob = model(chunk, SAMPLING_RATE).item()
|
| 189 |
speech_probs.append(speech_prob)
|
| 190 |
vad_iterator.reset_states() # reset model states after each audio
|
| 191 |
+
|
| 192 |
voice_idxs = np.where(np.array(speech_probs) >= 0.7)
|
| 193 |
+
print(len(np.where(np.array(speech_probs) >= 0.7))
|
| 194 |
if len(voice_idxs) == 0:
|
| 195 |
print("NO VOICE SEGMENTS DETECTED!")
|
| 196 |
try:
|
|
|
|
| 203 |
begin_seq = False
|
| 204 |
if voice_idxs[i+1] == voice_idxs[i]+1:
|
| 205 |
continue
|
| 206 |
+
|
| 207 |
start_time = float((start_idx*window_size_samples)/SAMPLING_RATE)
|
| 208 |
end_time = float((voice_idxs[i]*window_size_samples)/SAMPLING_RATE)
|
| 209 |
|
|
|
|
| 211 |
end_minutes = int(end_time)
|
| 212 |
start_seconds = (start_time - start_minutes) * 60
|
| 213 |
end_seconds = (end_time - end_minutes) * 60
|
| 214 |
+
|
| 215 |
+
print("modifying json data... \n")
|
| 216 |
data['vocal_times'] = {
|
| 217 |
"start_time": f"{start_minutes}.{start_seconds:.2f}",
|
| 218 |
"end_time": f"{end_minutes}.{end_seconds:.2f}"
|
|
|
|
| 226 |
print(f"An exception occurred: {e}")
|
| 227 |
|
| 228 |
with open(json_file, 'w') as f:
|
| 229 |
+
print("writing_to_json...")
|
| 230 |
json.dump(data, f, indent=4)
|
| 231 |
|
| 232 |
with gr.Blocks() as demo:
|