Xiaobo Xia

5 papers The University of Sydney

• Continual Multimodal Contrastive Learning
• L-MTP: Leap Multi-Token Prediction Beyond Adjacent Context for Large Language Models
• OpenOmni: Advancing Open-Source Omnimodal Large Language Models with Progressive Multimodal Alignment and Real-time Emotional Speech Synthesis
• UtilGen: Utility-Centric Generative Data Augmentation with Dual-Level Task Adaptation
• VCM: Vision Concept Modeling with Adaptive Vision Token Compression via Instruction Fine-Tuning