initinal

README.md

@@ -1,3 +1,6 @@
 | Attempt | #1    | #2    |
 | :---:   | :---: | :---: |
 | Seconds | 301   | 283   |
+
+Metrics on valentini dataset with baseline = {'PESQ': 1.5693, 'STOI': 0.9504}
+Metrics on valentini dataset with ideal denoising = {'PESQ': 1.9709, 'STOI': 0.9211}

denoisers/SpectralGating.py

@@ -4,7 +4,6 @@ import torchaudio
 
 
 class SpectralGating(torch.nn.Module):
-    """example: wav_noisy = '/media/public/datasets/denoising/DS_10283_2791/noisy_trainset_56spk_wav/p312_002.wav' """
     def __init__(self, rate=16000):
         super(SpectralGating, self).__init__()
         self.rate = rate

evaluation.py

@@ -15,8 +15,9 @@ MODELS = {
 
 
 
-def evaluate_on_dataset(model_name, dataset_path, dataset_type, ideal):
-    model = MODELS[model_name]()
+def evaluate_on_dataset(model_name, dataset_path, dataset_type):
+    if model_name is not None:
+        model = MODELS[model_name]()
     parser = PARSERS[dataset_type]
     clean_wavs, noisy_wavs = parser(dataset_path)
 
@@ -25,10 +26,11 @@ def evaluate_on_dataset(model_name, dataset_path, dataset_type, ideal):
     for clean_path, noisy_path in tqdm(zip(clean_wavs, noisy_wavs), total=len(clean_wavs)):
         clean_wav = load_wav(clean_path)
         noisy_wav = load_wav(noisy_path)
-        denoised_wav = model(noisy_wav)
-        if ideal:
+
+        if model_name is None:
             scores = metrics.calculate(noisy_wav, clean_wav)
         else:
+            denoised_wav = model(noisy_wav)
             scores = metrics.calculate(noisy_wav, denoised_wav)
 
         mean_scores['PESQ'] += scores['PESQ']
@@ -49,14 +51,12 @@ if __name__ == '__main__':
                         choices=['valentini'])
     parser.add_argument('--model_name', type=str,
                         choices=['baseline'])
-    parser.add_argument('--ideal', type=bool, default=False,
-                        help="Evaluate metrics on testing data with ideal denoising")
+
 
     args = parser.parse_args()
 
     mean_scores = evaluate_on_dataset(model_name=args.model_name,
                         dataset_path=args.dataset_path,
-                        dataset_type=args.dataset_type,
-                        ideal=args.ideal)
+                        dataset_type=args.dataset_type)
     print(f"Metrics on {args.dataset_type} dataset with "
           f"{args.model_name if args.model_name is not None else 'ideal denoising'} = {mean_scores}")