12 papers
UNISON: A Unified Framework for Learning from Scored Bags
Ofek Kirshenboim, Or Shkuri, Shani Finkelshtein +1
We introduce UNISON, a unified framework that bridges multiple instance learning and recommendation systems by treating both as instances of learning from scored bag collections of…
Object-Centric Conditioning for Visuomotor Flow Matching
Jijie Li, Xu Yang, Junhong Zou +4
Robot visuomotor policies are commonly formulated as autoregressive, diffusion-based, or more recently, flow matching models. Among them, Action-to-Action (A2A) flow matching impro…
Temporal Forcing: 4D Representation Alignment for Vision-Language-Action Models
Xingyu Ding, Yuzhong Zhao, Chunhai Zhao +3
Long-horizon robotic manipulation requires vision-language-action (VLA) models to track scene states and their evolution beyond the current observation. However, simply conditionin…
Mind-VLA: Instruction-Aware Spatial Representation Alignment for Vision-Language-Action Models
Xingyu Ding, Yuzhong Zhao, Yang Wu +4
Recent Vision-Language-Action (VLA) methods improve generalization by aligning their representations with 3D scene geometry. However, these methods are fundamentally instruction-ag…
ReST-KV: Robust KV Cache Eviction with Layer-wise Output Reconstruction and Spatial-Temporal Smoothing
Yongqi An, Chang Lu, Kuan Zhu +5
Large language models (LLMs) face growing challenges in efficient generative inference due to the increasing memory demands of Key-Value (KV) caches, especially for long sequences.…
PLUME: Latent Reasoning Based Universal Multimodal Embedding
Chenwei He, Xiangzhao Hao, Tianyu Yang +6
Universal multimodal embedding (UME) maps heterogeneous inputs into a shared retrieval space with a single model. Recent approaches improve UME by generating explicit chain-of-thou…
Listening with the Eyes: Benchmarking Egocentric Co-Speech Grounding across Space and Time
Weijie Zhou, Xuantang Xiong, Zhenlin Hu +6
In situated collaboration, speakers often use intentionally underspecified deictic commands (e.g., ``pass me \textit{that}''), whose referent becomes identifiable only by aligning…
ESearch-R1: Learning Cost-Aware MLLM Agents for Interactive Embodied Search via Reinforcement Learning
Weijie Zhou, Xuangtang Xiong, Ye Tian +8
Multimodal Large Language Models (MLLMs) have empowered embodied agents with remarkable capabilities in planning and reasoning. However, when facing ambiguous natural language inst…
PhysVLM-AVR: Active Visual Reasoning for Multimodal Large Language Models in Physical Environments
Weijie Zhou, Xuantang Xiong, Yi Peng +5
Visual reasoning in multimodal large language models (MLLMs) has primarily been studied in static, fully observable settings, limiting their effectiveness in real-world environment…
FOCUS: Fine-grained Optimization with Semantic Guided Understanding for Pedestrian Attributes Recognition
Hongyan An, Kuan Zhu, Xin He +4
Pedestrian attribute recognition (PAR) is a fundamental perception task in intelligent transportation and security. To tackle this fine-grained task, most existing methods focus on…
PhysVLM: Enabling Visual Language Models to Understand Robotic Physical Reachability
Weijie Zhou, Manli Tao, Chaoyang Zhao +4
Understanding the environment and a robot's physical reachability is crucial for task execution. While state-of-the-art vision-language models (VLMs) excel in environmental percept…
LightPlanner: Unleashing the Reasoning Capabilities of Lightweight Large Language Models in Task Planning
Weijie Zhou, Manli Tao, Chaoyang Zhao +3
In recent years, lightweight large language models (LLMs) have garnered significant attention in the robotics field due to their low computational resource requirements and suitabi…