Picture for Xiawu Zheng

Xiawu Zheng

QuoTA: Query-oriented Token Assignment via CoT Query Decouple for Long Video Comprehension

Add code
Mar 11, 2025
Viaarxiv icon

Dynamic Low-Rank Sparse Adaptation for Large Language Models

Add code
Feb 20, 2025
Viaarxiv icon

Determining Layer-wise Sparsity for Large Language Models Through a Theoretical Perspective

Add code
Feb 20, 2025
Viaarxiv icon

Towards Efficient Automatic Self-Pruning of Large Language Models

Add code
Feb 20, 2025
Viaarxiv icon

Training-free Anomaly Event Detection via LLM-guided Symbolic Pattern Discovery

Add code
Feb 09, 2025
Figure 1 for Training-free Anomaly Event Detection via LLM-guided Symbolic Pattern Discovery
Figure 2 for Training-free Anomaly Event Detection via LLM-guided Symbolic Pattern Discovery
Figure 3 for Training-free Anomaly Event Detection via LLM-guided Symbolic Pattern Discovery
Figure 4 for Training-free Anomaly Event Detection via LLM-guided Symbolic Pattern Discovery
Viaarxiv icon

Long-VITA: Scaling Large Multi-modal Models to 1 Million Tokens with Leading Short-Context Accuray

Add code
Feb 07, 2025
Viaarxiv icon

Solving the Catastrophic Forgetting Problem in Generalized Category Discovery

Add code
Jan 09, 2025
Viaarxiv icon

VITA-1.5: Towards GPT-4o Level Real-Time Vision and Speech Interaction

Add code
Jan 03, 2025
Viaarxiv icon

Video-RAG: Visually-aligned Retrieval-Augmented Long Video Comprehension

Add code
Nov 20, 2024
Figure 1 for Video-RAG: Visually-aligned Retrieval-Augmented Long Video Comprehension
Figure 2 for Video-RAG: Visually-aligned Retrieval-Augmented Long Video Comprehension
Figure 3 for Video-RAG: Visually-aligned Retrieval-Augmented Long Video Comprehension
Figure 4 for Video-RAG: Visually-aligned Retrieval-Augmented Long Video Comprehension
Viaarxiv icon

VITA: Towards Open-Source Interactive Omni Multimodal LLM

Add code
Aug 09, 2024
Figure 1 for VITA: Towards Open-Source Interactive Omni Multimodal LLM
Figure 2 for VITA: Towards Open-Source Interactive Omni Multimodal LLM
Figure 3 for VITA: Towards Open-Source Interactive Omni Multimodal LLM
Figure 4 for VITA: Towards Open-Source Interactive Omni Multimodal LLM
Viaarxiv icon