Hugging Face's logo

ccmusic-database
/
bel_canto

Audio Classification
English
music
art
Model card Files
xet
admin commited on
Commit
c31fa43
·
1 Parent(s): d6ddaa7

upd pic urls

Browse files
Files changed (1) hide show
  1. README.md +15 -4
README.md CHANGED
@@ -30,7 +30,18 @@ cd bel_canto
30
  ```
31
 
32
  ## Results
33
- A fine-tuning result of SqueezeNet on CQT:
 
 
 
 
 
 
 
 
 
 
 
34
  <style>
35
  #bel td {
36
  vertical-align: middle !important;
@@ -43,15 +54,15 @@ A fine-tuning result of SqueezeNet on CQT:
43
  <table id="bel">
44
  <tr>
45
  <th>Loss curve</th>
46
- <td><img src="https://www.modelscope.cn/models/ccmusic-database/bel_canto/resolve/master/squeezenet1_1_cqt_2024-07-30_04-38-35/loss.jpg"></td>
47
  </tr>
48
  <tr>
49
  <th>Training and validation accuracy</th>
50
- <td><img src="https://www.modelscope.cn/models/ccmusic-database/bel_canto/resolve/master/squeezenet1_1_cqt_2024-07-30_04-38-35/acc.jpg"></td>
51
  </tr>
52
  <tr>
53
  <th>Confusion matrix</th>
54
- <td><img src="https://www.modelscope.cn/models/ccmusic-database/bel_canto/resolve/master/squeezenet1_1_cqt_2024-07-30_04-38-35/mat.jpg"></td>
55
  </tr>
56
  </table>
57
 
 
30
  ```
31
 
32
  ## Results
33
+ | Backbone | Mel | CQT | Chroma |
34
+ | :-----------: | :-----------------------: | :-------: | :-------: |
35
+ | Swin-S | **0.928** | **0.936** | **0.787** |
36
+ | Swin-T | 0.906 | 0.863 | 0.731 |
37
+ | AlexNet | 0.919 | 0.920 | 0.746 |
38
+ | ConvNeXt-T | 0.895 | 0.925 | 0.714 |
39
+ | GoogleNet | [**0.948**](#best-result) | 0.921 | 0.739 |
40
+ | MNASNet1.3 | 0.931 | **0.931** | **0.765** |
41
+ | SqueezeNet1.1 | 0.923 | 0.914 | 0.685 |
42
+ | Average | 0.921 | 0.916 | 0.738 |
43
+
44
+ ### Best Result
45
  <style>
46
  #bel td {
47
  vertical-align: middle !important;
 
54
  <table id="bel">
55
  <tr>
56
  <th>Loss curve</th>
57
+ <td><img src="https://www.modelscope.cn/models/ccmusic-database/bel_canto/resolve/master/googlenet_mel_2024-07-30_00-51-26/loss.jpg"></td>
58
  </tr>
59
  <tr>
60
  <th>Training and validation accuracy</th>
61
+ <td><img src="https://www.modelscope.cn/models/ccmusic-database/bel_canto/resolve/master/googlenet_mel_2024-07-30_00-51-26/acc.jpg"></td>
62
  </tr>
63
  <tr>
64
  <th>Confusion matrix</th>
65
+ <td><img src="https://www.modelscope.cn/models/ccmusic-database/bel_canto/resolve/master/googlenet_mel_2024-07-30_00-51-26/mat.jpg"></td>
66
  </tr>
67
  </table>
68