Commit
•
866a8f3
1
Parent(s):
3371093
Update README.md
Browse files
README.md
CHANGED
|
@@ -22,6 +22,7 @@ tags:
|
|
| 22 |
- SER
|
| 23 |
- Transformer
|
| 24 |
- HuBERT
|
|
|
|
| 25 |
---
|
| 26 |
|
| 27 |
# **ExHuBERT: Enhancing HuBERT Through Block Extension and Fine-Tuning on 37 Emotion Datasets**
|
|
@@ -30,9 +31,7 @@ Authors: Shahin Amiriparian, Filip Packań, Maurice Gerczuk, Björn W. Schuller
|
|
| 30 |
Fine-tuned and backbone extended [**HuBERT Large**](https://huggingface.co/facebook/hubert-large-ls960-ft) on EmoSet++, comprising 37 datasets, totaling 150,907 samples and spanning a cumulative duration of 119.5 hours.
|
| 31 |
The model is expecting a 3 second long raw waveform resampled to 16 kHz. The original 6 Ouput classes are combinations of low/high arousal and negative/neutral/positive
|
| 32 |
valence.
|
| 33 |
-
Further details are available in the corresponding [**paper**](https://arxiv.org/)
|
| 34 |
-
|
| 35 |
-
**Note**: This model is for research purpose only.
|
| 36 |
|
| 37 |
### EmoSet++ subsets used for fine-tuning the model:
|
| 38 |
|
|
|
|
| 22 |
- SER
|
| 23 |
- Transformer
|
| 24 |
- HuBERT
|
| 25 |
+
- Affective Computing
|
| 26 |
---
|
| 27 |
|
| 28 |
# **ExHuBERT: Enhancing HuBERT Through Block Extension and Fine-Tuning on 37 Emotion Datasets**
|
|
|
|
| 31 |
Fine-tuned and backbone extended [**HuBERT Large**](https://huggingface.co/facebook/hubert-large-ls960-ft) on EmoSet++, comprising 37 datasets, totaling 150,907 samples and spanning a cumulative duration of 119.5 hours.
|
| 32 |
The model is expecting a 3 second long raw waveform resampled to 16 kHz. The original 6 Ouput classes are combinations of low/high arousal and negative/neutral/positive
|
| 33 |
valence.
|
| 34 |
+
Further details are available in the corresponding [**paper**](https://arxiv.org/).
|
|
|
|
|
|
|
| 35 |
|
| 36 |
### EmoSet++ subsets used for fine-tuning the model:
|
| 37 |
|