Heng-Jui Chang

vectominist

·

https://people.csail.mit.edu/hengjui/

AI & ML interests

Speech Processing, Multimodal Learning, Self-supervised Learning

Recent Activity

authored a paper 24 days ago

End-to-end Whispered Speech Recognition with Frequency-weighted Approaches and Pseudo Whisper Pre-training

authored a paper 24 days ago

USAD 2.0: Scaling Representation Distillation for Universal Audio Understanding

updated a model 26 days ago

MIT-SLS/USAD-Large

View all activity

Organizations

upvoted a paper 27 days ago

USAD 2.0: Scaling Representation Distillation for Universal Audio Understanding

Paper • 2606.06444 • Published 28 days ago • 3

upvoted a collection 6 months ago

perception-encoder-audio-visual

9 items • Updated Dec 16, 2025 • 29

upvoted a collection 9 months ago

Meta CLIP 1

Scaling CLIP data with transparent training distribution from an end-to-end pipeline. • 7 items • Updated Nov 24, 2025 • 23

upvoted a paper about 1 year ago

USAD: Universal Speech and Audio Representation via Distillation

Paper • 2506.18843 • Published Jun 23, 2025 • 13

upvoted a paper over 1 year ago

SelfCite: Self-Supervised Alignment for Context Attribution in Large Language Models

Paper • 2502.09604 • Published Feb 13, 2025 • 36