🧪 Test?View on arXiv
Whispering Wisely: Smart Subset Selection for ASR Fine-Tuning
Author1, Author2, Author3, Author4, Author5
fine-tuningASRsubset selectionembedding
2610.32997
Builder Relevance
1h ago80%
Abstract
This paper presents an embedding-based subset selection approach for fine-tuning ASR models, demonstrating improved performance with smaller training sets.
Reality Card
Core Claim
The embedding-based subset selection approach using a greedy k-center algorithm improves ASR fine-tuning effectiveness, particularly with smaller training subsets.
Method / Result
k-center selection improves performance over random sampling, especially at smaller subset sizes.
Limitations
The approach has only been evaluated on Hebrew and Danish, limiting generalizability to other languages or domains.
Paper to code
Verified implementation resources so builders can test the paper’s claims instead of stopping at the abstract.
No verified implementation link has been attached yet. AIBuzzHub will keep this panel separate from unverified search results.