10 papers
StateTrace: An Object-Centric Framework for Hidden-State Spatiotemporal Reasoning in Long Videos
Yu Han, Wenhao Li, Yichao Cao +4
Existing VLMs have achieved strong performance in video understanding, yet they struggle with long-video spatiotemporal reasoning when target objects become invisible, often mistak…
Do All Individual Layers Help? An Empirical Study of Task-Interfering Layers in Vision-Language Models
Zhiming Liu, Yujie Wei, Lei Feng +5
Current VLMs have demonstrated capabilities across a wide range of multimodal tasks. Typically, in a pretrained VLM, all layers are engaged by default to make predictions on downst…
Learning to Accelerate Vision-Language-Action Models through Adaptive Visual Token Caching
Yujie Wei, Jiahan Fan, Jiyu Guo +6
Vision-Language-Action (VLA) models have demonstrated remarkable generalization capabilities in robotic manipulation tasks, yet their substantial computational overhead remains a c…
ConLA: Contrastive Latent Action Learning from Human Videos for Robotic Manipulation
Weisheng Dai, Kai Lan, Jianyi Zhou +5
Vision-Language-Action (VLA) models achieve preliminary generalization through pretraining on large scale robot teleoperation datasets. However, acquiring datasets that comprehensi…
APEX: A Decoupled Memory-based Explorer for Asynchronous Aerial Object Goal Navigation
Daoxuan Zhang, Ping Chen, Xiaobo Xia +4
Aerial Object Goal Navigation, a challenging frontier in Embodied AI, requires an Unmanned Aerial Vehicle (UAV) agent to autonomously explore, reason, and identify a specific targe…
Inject Once Survive Later: Backdooring Vision-Language-Action Models to Persist Through Downstream Fine-tuning
Jianyi Zhou, Yujie Wei, Ruichen Zhen +5
Vision-Language-Action (VLA) models have become foundational to modern embodied AI systems. By integrating visual perception, language understanding, and action planning, they enab…
Calibrated Multimodal Representation Learning with Missing Modalities
Xiaohao Liu, Xiaobo Xia, Jiaheng Wei +4
Multimodal representation learning harmonizes distinct modalities by aligning them into a unified latent space. Recent research generalizes traditional cross-modal alignment to pro…
UtilGen: Utility-Centric Generative Data Augmentation with Dual-Level Task Adaptation
Jiyu Guo, Shuo Yang, Yiming Huang +6
Data augmentation using generative models has emerged as a powerful paradigm for enhancing performance in computer vision tasks. However, most existing augmentation approaches prim…
Logic Unseen: Revealing the Logical Blindspots of Vision-Language Models
Yuchen Zhou, Jiayu Tang, Shuo Yang +6
Vision-Language Models (VLMs), exemplified by CLIP, have emerged as foundational for multimodal intelligence. However, their capacity for logical understanding remains significantl…
L-MTP: Leap Multi-Token Prediction Beyond Adjacent Context for Large Language Models
Xiaohao Liu, Xiaobo Xia, Weixiang Zhao +6
Large language models (LLMs) have achieved notable progress. Despite their success, next-token prediction (NTP), the dominant method for LLM training and inference, is constrained…