We further introduce an ASR-anchored supervision method that projects word-level turn annotations onto the frame-level ASR token timeline

We design a unified turn state label set that supports interruption, turn completion, backchannel detection

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it

browse 1 citing papers

representative citing papers

X2-Turn: Frame-Synchronous Dual-Head Modeling for Joint Streaming ASR and Turn State Prediction

cs.CL · 2026-08-11

citing papers explorer

Showing 1 of 1 citing paper.

X2-Turn: Frame-Synchronous Dual-Head Modeling for Joint Streaming ASR and Turn State Prediction cs.CL · 2026-08-11 · unreviewed · ref 3