Update README.md
Browse files
README.md
CHANGED
|
@@ -7,8 +7,15 @@ tags: []
|
|
| 7 |
|
| 8 |
ConvNext trained on the XCL dataset from BirdSet, covering 9736 bird species from Xeno-Canto. Please refer to the [BirdSet Paper](https://arxiv.org/pdf/2403.10380) and the
|
| 9 |
[BirdSet Repository](https://github.com/DBD-research-group/BirdSet/tree/main) for further information.
|
| 10 |
-
The model is trained on 5-second clips of bird vocalizations.
|
| 11 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
|
| 13 |
### Model Details
|
| 14 |
ConvNeXT is a pure convolutional model (ConvNet), inspired by the design of Vision Transformers, that claims to outperform them.
|
|
|
|
| 7 |
|
| 8 |
ConvNext trained on the XCL dataset from BirdSet, covering 9736 bird species from Xeno-Canto. Please refer to the [BirdSet Paper](https://arxiv.org/pdf/2403.10380) and the
|
| 9 |
[BirdSet Repository](https://github.com/DBD-research-group/BirdSet/tree/main) for further information.
|
|
|
|
| 10 |
|
| 11 |
+
- The model is trained on 5-second clips of bird vocalizations.
|
| 12 |
+
- num_channels: 1
|
| 13 |
+
- pretrained checkpoint: facebook/convnext-base-224-22k
|
| 14 |
+
- sampling_rate: 32_000
|
| 15 |
+
- normalize spectrogram: mean: -4.268, std: 4.569 (from esc-50)
|
| 16 |
+
- spectrogram: n_fft: 1024, hop_length: 320, power: 2.0
|
| 17 |
+
- melscale: n_mels: 128, n_stft: 513
|
| 18 |
+
- dbscale: top_db: 80
|
| 19 |
|
| 20 |
### Model Details
|
| 21 |
ConvNeXT is a pure convolutional model (ConvNet), inspired by the design of Vision Transformers, that claims to outperform them.
|