DOA: Training-Free Decoder-Only Attention Policy for Long-Form Simultaneous Translation with SpeechLLMs Paper • 2605.31432 • Published 27 days ago • 9
SimulU: Training-free Policy for Long-form Simultaneous Speech-to-Speech Translation Paper • 2603.16924 • Published Mar 11 • 16
Simulstream: Open-Source Toolkit for Evaluation and Demonstration of Streaming Speech-to-Text Translation Systems Paper • 2512.17648 • Published Dec 19, 2025 • 5
How to Evaluate Speech Translation with Source-Aware Neural MT Metrics Paper • 2511.03295 • Published Nov 5, 2025 • 4
Cross-Attention is Half Explanation in Speech-to-Text Models Paper • 2509.18010 • Published Sep 22, 2025 • 6
Cross-Attention is Half Explanation in Speech-to-Text Models Paper • 2509.18010 • Published Sep 22, 2025 • 6