Embedding-based subset selection improves ASR Fine-Tuning on hebrew and danish
Embedding-based subset selection uses Whisper’s pretrained encoder to map audio samples into a fixed‑dimensional embedding space. A greedy k‑center algorithm then selects diverse, representative subsets for fine‑tuning. The authors evaluated the approach on Hebrew and Danish, finding that k‑center subsets outperform random sampling, especially at smaller subset sizes. The method also captures…
1 source primary source
Key points
- Embedding-based subset selection uses Whisper encoder embeddings and greedy k-center to pick diverse training samples.
- Tested on Hebrew and Danish, k-center subsets outperform random sampling, especially with small subset sizes.
- The method yields broader lexical diversity, leading to better ASR fine-tuning performance.
Read the original at arXiv cs.CL · by Roi Yozevitch, Or Haim Anidjar primary sourceOpen source ↗
Topics · follow one to build your own front page
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Research
All →- SLICE-Mem cuts LLM memory costs by 99.5% in new framework · 1 src
- Researchers propose RSD metric to measure LLM agent consistency · 1 src
- Researchers propose tracequery to cut diffusion LLM inference costs · 1 src
- Researchers propose knowledge-guided LLM for 6G service orchestration · 1 src
- Arxiv paper introduces Boa Compressor for Linear-Time context compression · 1 src
Comments
via GitHub Discussions