habla-liaa/ser-with-w2v2

Official implementation of INTERSPEECH 2021 paper 'Emotion Recognition from Speech Using Wav2vec 2.0 Embeddings'

37
/ 100
Emerging

Leverages pretrained Wav2vec 2.0 encodings as fixed speech representations, combining encoder outputs with transformer layer embeddings through fusion strategies to extract emotion-relevant features. Evaluated on RAVDESS and IEMOCAP datasets using 5-fold cross-validation with multiple random seeds, providing pretrained checkpoints for both datasets alongside a reproducible experimental pipeline via YAML configuration and shell scripts.

140 stars. No commits in the last 6 months.

No License Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 8 / 25
Community 19 / 25

How are scores calculated?

Stars

140

Forks

25

Language

Jupyter Notebook

License

Last pushed

Jan 06, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/habla-liaa/ser-with-w2v2"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.