tiantiaf commited on
Commit
8db0ea8
·
verified ·
1 Parent(s): 0b29812

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +31 -1
README.md CHANGED
@@ -2,8 +2,38 @@
2
  tags:
3
  - model_hub_mixin
4
  - pytorch_model_hub_mixin
 
 
 
 
 
 
 
 
5
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
6
 
7
- This model has been pushed to the Hub using the [PytorchModelHubMixin](https://huggingface.co/docs/huggingface_hub/package_reference/mixins#huggingface_hub.PyTorchModelHubMixin) integration:
8
  - Library: https://github.com/tiantiaf0627/vox-profile-release
9
  - Docs: [More Information Needed]
 
2
  tags:
3
  - model_hub_mixin
4
  - pytorch_model_hub_mixin
5
+ license: apache-2.0
6
+ language:
7
+ - en
8
+ metrics:
9
+ - accuracy
10
+ base_model:
11
+ - openai/whisper-large-v3
12
+ pipeline_tag: audio-classification
13
  ---
14
+ # Whisper-Large V3 for Categorical Emotion Classification
15
+
16
+ # Model Description
17
+ This model includes the implementation of categorical emotion classification described in Vox-Profile: A Speech Foundation Model Benchmark for Characterizing Diverse Speaker and Speech Traits (https://arxiv.org/pdf/2505.14648)
18
+
19
+ The training pipeline used is also the top performing solution (SAILER) in INTERSPEECH 2025 - Speech Emotion Challenge (https://lab-msp.com/MSP-Podcast_Competition/IS2025/). Note that we did not use all the augmentation and and did not use the transcript to make the model simple but still effective.
20
+ We use the MSP-Podcast data for training this model. Noting that the model might be sensitive to content information in making the emotion prediction.
21
+
22
+
23
+ The included emotions are:
24
+ <pre>
25
+ [
26
+ 'Anger',
27
+ 'Contempt',
28
+ 'Disgust',
29
+ 'Fear',
30
+ 'Happiness',
31
+ 'Neutral',
32
+ 'Sadness',
33
+ 'Surprise',
34
+ 'Other'
35
+ ]
36
+ </pre>
37
 
 
38
  - Library: https://github.com/tiantiaf0627/vox-profile-release
39
  - Docs: [More Information Needed]