TY  - RPRT
TI  - VIB-AVSR: Variational Information Bottleneck for Noise-Robust LLM-Based Audio-Visual Speech Recognition
AU  - Piyush Arora
AU  - Navlika Singh
AU  - Umberto Cappellazzo
AU  - Stavros Petridis
AU  - Maja Pantic
PY  - 2026
UR  - https://arxiv.org/abs/2606.29632
ID  - 2606.29632
ER  -