Publications (7)
RainDancer: RGB-Event Video Deraining with Rain-Oriented Spiking Dynamics
Kui Jiang, Runzhe Li, Zhaocheng Yu +3
RainDancer is a video deraining framework that jointly processes RGB frames and event-camera data, first decomposing rain and background within each modality and then fusing them u…
Neural-Symbolic VideoQA: Learning Compositional Spatio-Temporal Reasoning for Real-world Video Question Answering
Lili Liang, Guanglu Sun, Jin Qiu +1
Compositional spatio-temporal reasoning poses a significant challenge in the field of video question answering (VideoQA). Existing approaches struggle to establish effective symbol…
Analyzing Reasoning Consistency in Large Multimodal Models under Cross-Modal Conflicts
Zhihao Zhu, Jiafeng Liang, Shixin Jiang +5
Large Multimodal Models (LMMs) have demonstrated impressive capabilities in video reasoning via Chain-of-Thought (CoT). However, the robustness of their reasoning chains remains qu…
Semantics and Content Matter: Towards Multi-Prior Hierarchical Mamba for Image Deraining
Zhaocheng Yu, Kui Jiang, Junjun Jiang +3
Rain significantly degrades the performance of computer vision systems, particularly in applications like autonomous driving and video surveillance. While existing deraining method…
Leveraging Static Relationships for Intra-Type and Inter-Type Message Passing in Video Question Answering
Lili Liang, Guanglu Sun
Video Question Answering (VideoQA) is an important research direction in the field of artificial intelligence, enabling machines to understand video content and perform reasoning a…
Derain-Agent: A Plug-and-Play Agent Framework for Rainy Image Restoration
Zhaocheng Yu, Xiang Chen, Runzhe Li +4
While deep learning has advanced single-image deraining, existing models suffer from a fundamental limitation: they employ a static inference paradigm that fails to adapt to the co…
Unbiased Scene Graph Generation by Type-Aware Message Passing on Heterogeneous and Dual Graphs
Guanglu Sun, Jin Qiu, Lili Liang
Although great progress has been made in the research of unbiased scene graph generation, issues still hinder improving the predictive performance of both head and tail classes. An…