AutoTerm-SST: Adaptive Terminology Memory with Zero Session-Time Setup for Streaming Speech Translation

Not specified in the provided content

streaming translationterminology managementreal-time processing

2610.26865

Builder Relevance

80%

2h ago

Abstract

AutoTerm-SST presents an interactive system for streaming speech translation that eliminates the need for session-time glossary uploads and domain selection.

Reality Card

Core Claim

AutoTerm-SST achieves comparable terminology accuracy and BLEU scores to existing systems while requiring no domain input, improving terminology accuracy by 3.9 points and BLEU by 2.30.

Method / Result

Terminology accuracy of 88.3% and mWER-resegmented BLEU of 45.09.

Limitations

The paper does not specify the authors or provide detailed reproducibility information.

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.

← Back to all papers