Liang Zheng
2 papers · Latest:
Natural Language Processing
WARDEN: Endangered Indigenous Language Transcription and Translation with 6 Hours of Training Data
WARDEN is a novel two-stage system that transcribes and translates endangered Wardaman to English using only 6 hours of audio, outperforming larger models.
2605.13846
Computer VisionLeapAlign: Post-Training Flow Matching Models at Any Generation Step by Building Two-Step Trajectories
LeapAlign fine-tunes flow matching models for preference alignment by using two-step trajectories, enabling efficient, stable direct gradient updates at any generation step.
2604.15311
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.