-
Notifications
You must be signed in to change notification settings - Fork 0
/
old-main.py
279 lines (230 loc) · 9.49 KB
/
old-main.py
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
import streamlit as st
import os
import librosa
import numpy as np
import pywt
from scipy import signal
import wave
import array
import math
import tempfile
import boto3
import uuid
import torch
from torch import nn
import torchaudio
import torchaudio.transforms as T
from dotenv import load_dotenv
load_dotenv()
AWS_ACCESS_KEY_ID = os.getenv('AWS_ACCESS_KEY_ID')
AWS_SECRET_ACCESS_KEY = os.getenv('AWS_SECRET_ACCESS_KEY')
REGION_NAME = os.getenv('REGION_NAME')
BUCKET_NAME = os.getenv('BUCKET_NAME')
s3 = boto3.client('s3')
def upload_to_s3(file_content, file_name):
bucket_name = BUCKET_NAME
folder_name = str(uuid.uuid4()) # Generate a random folder name
s3_path = f"{folder_name}/{file_name}"
try:
s3.put_object(Bucket=bucket_name, Key=s3_path, Body=file_content)
return f"s3://{bucket_name}/{s3_path}"
except Exception as e:
st.error(f"Error uploading audio file: {e}")
return None
def download_from_s3(s3_uri):
s3 = boto3.client('s3')
bucket_name = s3_uri.split('/')[2]
s3_path = '/'.join(s3_uri.split('/')[3:])
try:
with tempfile.NamedTemporaryFile(delete=False) as tmp_file:
s3.download_fileobj(bucket_name, s3_path, tmp_file)
return tmp_file.name
except Exception as e:
st.error(f"Error downloading from S3: {e}")
return None
def peak_detect(data):
max_val = np.amax(abs(data))
peak_ndx = np.where(data == max_val)
if len(peak_ndx[0]) == 0: # if nothing found then the max must be negative
peak_ndx = np.where(data == -max_val)
return peak_ndx[0]
def wavelet_bpm_detector(audio_file_path):
try:
samps, fs = read_wav(audio_file_path)
data = samps
cA = []
cD = []
correl = []
cD_sum = []
levels = 4
max_decimation = 2 ** (levels - 1)
min_ndx = math.floor(60.0 / 220 * (fs / max_decimation))
max_ndx = math.floor(60.0 / 40 * (fs / max_decimation))
for loop in range(0, levels):
cD = []
if loop == 0:
[cA, cD] = pywt.dwt(data, "db4")
cD_minlen = len(cD) // max_decimation + 1
cD_sum = np.zeros(math.floor(cD_minlen))
else:
[cA, cD] = pywt.dwt(cA, "db4")
cD = signal.lfilter([0.01], [1 - 0.99], cD)
cD = abs(cD[:: (2 ** (levels - loop - 1))])
cD = cD - np.mean(cD)
cD_sum = cD[0 : math.floor(cD_minlen)] + cD_sum
if [b for b in cA if b != 0.0] == []:
return None
cA = signal.lfilter([0.01], [1 - 0.99], cA)
cA = abs(cA)
cA = cA - np.mean(cA)
cD_sum = cA[0 : math.floor(cD_minlen)] + cD_sum
correl = np.correlate(cD_sum, cD_sum, "full")
midpoint = len(correl) // 2
correl_midpoint_tmp = correl[midpoint:]
peak_ndx = peak_detect(correl_midpoint_tmp[min_ndx:max_ndx])
if len(peak_ndx) > 1:
return None
peak_ndx_adjusted = peak_ndx[0] + min_ndx
bpm = 60.0 / peak_ndx_adjusted * (fs / max_decimation)
return int(round(bpm))
except Exception as e:
print(f"Error in wavelet BPM detection: {e}")
return None
def read_wav(filename):
try:
y, sr = librosa.load(filename, sr=None)
return y, sr
except Exception as e:
print(f"Error reading audio file: {e}")
return None, None
class BPMPredictor(nn.Module):
def __init__(self, input_size):
super(BPMPredictor, self).__init__()
self.fc1 = nn.Linear(input_size, 256)
self.bn1 = nn.BatchNorm1d(256)
self.fc2 = nn.Linear(256, 128)
self.bn2 = nn.BatchNorm1d(128)
self.fc3 = nn.Linear(128, 64)
self.bn3 = nn.BatchNorm1d(64)
self.fc4 = nn.Linear(64, 1)
self.dropout = nn.Dropout(0.3)
self.relu = nn.LeakyReLU(0.1)
def forward(self, x):
x = x.view(x.size(0), -1) # Flatten the input
x = self.dropout(self.relu(self.bn1(self.fc1(x))))
x = self.dropout(self.relu(self.bn2(self.fc2(x))))
x = self.dropout(self.relu(self.bn3(self.fc3(x))))
x = self.fc4(x)
return x
def extract_audio_features(file_path, device):
y, sr = librosa.load(file_path, sr=None)
# Extract features
mfcc = librosa.feature.mfcc(y=y, sr=sr, n_mfcc=13)
mel = librosa.feature.melspectrogram(y=y, sr=sr, n_mels=64)
chroma = librosa.feature.chroma_stft(y=y, sr=sr)
# Compute statistics
features = np.concatenate([
np.mean(mfcc, axis=1),
np.mean(mel, axis=1),
np.mean(chroma, axis=1),
])
return torch.FloatTensor(features).to(device)
def detect_bpm(audio_file):
try:
# Existing methods
y, sr = librosa.load(audio_file)
onset_env = librosa.onset.onset_strength(y=y, sr=sr)
tempo, beats = librosa.beat.beat_track(onset_envelope=onset_env, sr=sr)
tempo_from_tempogram = librosa.feature.tempo(onset_envelope=onset_env, sr=sr)[0]
# New wavelet-based method
wavelet_bpm = wavelet_bpm_detector(audio_file)
# Load the trained model
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
input_size = 13 + 64 + 12 # 13 (MFCC) + 64 (Mel) + 12 (Chroma)
model = BPMPredictor(input_size)
model_path = os.path.join('model-train', 'best_model.pth')
model.load_state_dict(torch.load(model_path, map_location=device, weights_only=True))
model.to(device)
model.eval()
# Extract features and predict BPM
features = extract_audio_features(audio_file, device)
with torch.no_grad():
model_bpm = model(features.unsqueeze(0)).item()
# Combine all methods
all_tempos = [tempo, tempo_from_tempogram, model_bpm]
if wavelet_bpm is not None:
all_tempos.append(wavelet_bpm)
# Check for half or double tempo
potential_tempos = []
for t in all_tempos:
if isinstance(t, np.ndarray):
t = t.item()
potential_tempos.extend([t/2, t, t*2])
potential_tempos = [t for t in potential_tempos if 60 <= t <= 200]
if not potential_tempos:
return int(round(np.mean(all_tempos)))
# Choose the tempo that best matches the onset strength
best_tempo = max(potential_tempos, key=lambda t: onset_strength_at_tempo(onset_env, sr, t))
return int(round(best_tempo))
except Exception as e:
st.error(f"Error detecting BPM: {str(e)}")
return None
def verify_and_adjust_bpm(audio, detected_bpm):
duration_ms = len(audio)
beat_duration_ms = (60 / detected_bpm) * 1000
expected_beats = duration_ms / beat_duration_ms
rounded_beats = round(expected_beats)
if abs(expected_beats - rounded_beats) < 0.1: # 10% tolerance
return detected_bpm
# If not matching, try adjusting BPM
adjusted_bpm = (detected_bpm * rounded_beats) / expected_beats
# Check if doubling or halving the BPM would be more accurate
if abs(adjusted_bpm - detected_bpm * 2) < abs(adjusted_bpm - detected_bpm):
return int(round(detected_bpm * 2))
elif abs(adjusted_bpm - detected_bpm / 2) < abs(adjusted_bpm - detected_bpm):
return int(round(detected_bpm / 2))
return int(round(adjusted_bpm))
def onset_strength_at_tempo(onset_env, sr, tempo):
# Helper function to calculate onset strength at a given tempo
tempo_period = 60.0 / tempo
hop_length = 512 # This should match the hop_length used in librosa.onset.onset_strength
beats = librosa.util.fix_frames(np.arange(0, len(onset_env), tempo_period * sr / hop_length))
return np.mean(onset_env[beats])
# Streamlit app
def main():
st.title("Audio BPM Detector")
st.write("Upload an audio file (MP3, FLAC, or WAV) to detect its BPM.")
uploaded_file = st.file_uploader("Choose an audio file", type=["mp3", "flac", "wav"])
if uploaded_file is not None:
file_details = {"FileName": uploaded_file.name, "FileType": uploaded_file.type, "FileSize": uploaded_file.size}
st.write(file_details)
# Check file size
if uploaded_file.size > 100 * 1024 * 1024: # 100 MB
st.error("File size exceeds the maximum limit of 100 MB.")
else:
# Upload to S3
s3_uri = upload_to_s3(uploaded_file.getvalue(), uploaded_file.name)
if s3_uri:
st.info(f"File uploaded to cloud")
st.info(f"Running detection on file...")
# Download from S3
tmp_file_path = download_from_s3(s3_uri)
if tmp_file_path:
try:
# Detect BPM
detected_bpm = detect_bpm(tmp_file_path)
if detected_bpm:
st.success(f"Detected BPM: {detected_bpm}")
else:
st.warning("Unable to detect BPM. Please try a different audio file.")
except Exception as e:
st.error(f"An error occurred: {e}")
finally:
# Clean up the temporary file
os.unlink(tmp_file_path)
else:
st.error("Failed to download file from S3.")
else:
st.error("Failed to upload file to S3.")
if __name__ == "__main__":
main()