8 papers
ViTexQA: A Multi-Frame Temporal Perception Dataset for Video Text Question Answering
Zhentao Guo, Chen Duan, Tongkun Guan +3
Despite remarkable progress in multimodal understanding, current MLLMs still exhibit limitations in video text understanding, particularly when semantics emerge through the integra…
PositionOCR: Augmenting Positional Awareness in Multi-Modal Models via Hybrid Specialist Integration
Chen Duan, Zhentao Guo, Pei Fu +3
In recent years, Multi-modal Large Language Models (MLLMs) have achieved strong performance in OCR-centric Visual Question Answering (VQA) tasks, illustrating their capability to p…
Driving with DINO: Vision Foundation Features as a Unified Bridge for Sim-to-Real Generation in Autonomous Driving
Xuyang Chen, Conglang Zhang, Chuanheng Fu +10
Driven by the emergence of Controllable Video Diffusion, existing Sim2Real methods for autonomous driving video generation typically rely on explicit intermediate representations t…
Context Forcing: Consistent Autoregressive Video Generation with Long Context
Shuo Chen, Cong Wei, Sun Sun +5
Recent approaches to real-time long video generation typically employ streaming tuning strategies, attempting to train a long-context student using a short-context (memoryless) tea…
MeSS: City Mesh-Guided Outdoor Scene Generation with Cross-View Consistent Diffusion
Xuyang Chen, Zhijun Zhai, Kaixuan Zhou +9
Mesh models have become increasingly accessible for numerous cities; however, the lack of realistic textures restricts their application in virtual urban navigation and autonomous…
A Token-level Text Image Foundation Model for Document Understanding
Tongkun Guan, Zining Wang, Pei Fu +9
In recent years, general visual foundation models (VFMs) have witnessed increasing adoption, particularly as image encoders for popular multi-modal large language models (MLLMs). H…