SER and audio classification using both a Wav2Vec2 based model and an ASR->Bert pipeline, as well as utilizing a multimodal late-fusion model
1
stars
4
commits
Python
primary language
Jul 4, 2023
updated
README
Wav2Vec2Stats
This is my official entry for the ACM ComPaRe 2023 challenge, performing Audio classificaion and SER using a Wav2Vec2-based model. Run asr.py to generate transcripts for the respective datasets, wav2vec2.py to run Wav2Vec2Stats, bert.py for the rudimentary BERT-based classifier on the transcripts, and evaluate.ipynb to generate the submission CSVs
Contributors
4 commits
Similar repos
Languages
Python
73.9%
Jupyter Notebook
26.1%
SER and audio classification using both a Wav2Vec2 based model and an ASR->Bert pipeline, as well as utilizing a multimodal late-fusion model
1
stars
4
commits
Python
primary language
Jul 4, 2023
updated
README
Wav2Vec2Stats
This is my official entry for the ACM ComPaRe 2023 challenge, performing Audio classificaion and SER using a Wav2Vec2-based model. Run asr.py to generate transcripts for the respective datasets, wav2vec2.py to run Wav2Vec2Stats, bert.py for the rudimentary BERT-based classifier on the transcripts, and evaluate.ipynb to generate the submission CSVs
Similar repos
Contributors
4 commits
Languages
Python
73.9%
Jupyter Notebook
26.1%