7 papers
Many Preferences, Few Policies: Towards Scalable Language Model Personalization
Cheol Woo Kim, Jai Moondra, Roozbeh Nahavandi +3
The holy grail of LLM personalization is a single LLM for each user, perfectly aligned with that user's preferences. However, maintaining a separate LLM per user is impractical due…
Do Audio LLMs Really LISTEN, or Just Transcribe? Measuring Lexical vs. Acoustic Emotion Cues Reliance
Jingyi Chen, Zhimeng Guo, Jiyun Chun +3
Understanding emotion from speech requires sensitivity to both lexical and acoustic cues. However, it remains unclear whether large audio language models (LALMs) genuinely process…
Fine-Tuning Text-to-Speech Diffusion Models Using Reinforcement Learning with Human Feedback
Jingyi Chen, Ju Seung Byun, Micha Elsner +2
Diffusion models produce high-fidelity speech but are inefficient for real-time use due to long denoising steps and challenges in modeling intonation and rhythm. To improve this, w…
Symmetric Reinforcement Learning Loss for Robust Learning on Diverse Tasks and Model Scales
Ju-Seung Byun, Andrew Perrault
Reinforcement learning (RL) training is inherently unstable due to factors such as moving targets and high gradient variance. Reinforcement Learning from Human Feedback (RLHF) and…
DLPO: Diffusion Model Loss-Guided Reinforcement Learning for Fine-Tuning Text-to-Speech Diffusion Models
Jingyi Chen, Ju-Seung Byun, Micha Elsner +1
Recent advancements in generative models have sparked a significant interest within the machine learning community. Particularly, diffusion models have demonstrated remarkable capa…
Using RLHF to align speech enhancement approaches to mean-opinion quality scores
Anurag Kumar, Andrew Perrault, Donald S. Williamson
Objective speech quality measures are typically used to assess speech enhancement algorithms, but it has been shown that they are sub-optimal as learning objectives because they do…