2 citations · 2 across the 4 of their papers we have counts for
6 papers
DuplexAct: How Interactive Are Frontier Full-Duplex Spoken Language Models?
Jiaheng Dong, Shihong Tan, Gongping Huang +3
End-to-end (E2E) full-duplex spoken language models (FD-SLMs) enable simutanous listening and speaking continously, requiring models to take appropriate interaction actions as conv…
Listen Then Reason: Perception-Grounded Test-Time Reinforcement Learning for Large Audio-Language Models
Jiaheng Dong, Xiaofeng Yu, Jean Honorio +3
Large audio-language models (LALMs) are increasingly used for a broader range of audio reasoning tasks. These models typically incorporate audio representations into a large langua…
Disentangling Reasoning in Large Audio-Language Models for Ambiguous Emotion Prediction
Xiaofeng Yu, Jiaheng Dong, Jean Honorio +3
Speech emotion recognition plays an important role in various applications. However, most existing approaches predict a single emotion label, oversimplifying the inherently ambiguo…
NEO: No-Optimization Test-Time Adaptation through Latent Re-Centering
Alexander Murphy, Michal Danilowski, Soumyajit Chatterjee +1
Test-Time Adaptation (TTA) methods are often computationally expensive, require a large amount of data for effective adaptation, or are brittle to hyperparameters. Based on a theor…
E-BATS: Efficient Backpropagation-Free Test-Time Adaptation for Speech Foundation Models
Jiaheng Dong, Hong Jia, Soumyajit Chatterjee +3
Speech Foundation Models encounter significant performance degradation when deployed in real-world scenarios involving acoustic domain shifts, such as background noise and speaker…
BoTTA: Benchmarking on-device Test Time Adaptation
Michal Danilowski, Soumyajit Chatterjee, Abhirup Ghosh
The performance of deep learning models depends heavily on test samples at runtime, and shifts from the training data distribution can significantly reduce accuracy. Test-time adap…