Create README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,20 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: mit
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
pipeline_tag: audio-classification
|
| 6 |
+
tags:
|
| 7 |
+
- pytorch
|
| 8 |
+
- wavlm
|
| 9 |
+
- msp-podcast
|
| 10 |
+
- emotion-recognition
|
| 11 |
+
- audio
|
| 12 |
+
- speech
|
| 13 |
+
- valence
|
| 14 |
+
- arousal
|
| 15 |
+
- dominance
|
| 16 |
+
- lucas
|
| 17 |
+
- speech-emotion-recognition
|
| 18 |
+
---
|
| 19 |
+
The model is a recreation of [3loi/SER-Odyssey-Baseline-WavLM-Multi-Attributes](https://huggingface.co/3loi/SER-Odyssey-Baseline-WavLM-Multi-Attributes) for direct implementation in torch, with class definition and feed forward method. This model was recreated with the hopes of greater flexibilty of control, training/fine-tuning of model. The model was trained on the same [MSP-Podcast](https://ecs.utdallas.edu/research/researchlabs/msp-lab/MSP-Podcast.html) dataset as the original, but a different smaller subset was used. Thesubset is evenly distributed across gender and emotion category.
|
| 20 |
+
This model is therefore a multi-attributed based model which predict arousal, dominance and valence. However, unlike the original model, I just kept the original attribute score range of 0...7 (the range the dataset follows). I will provide the
|