8 citations · 23 across the 45 of their papers we have counts for
53 papers
Understanding and Mitigating Token-Pruning-Induced Vulnerabilities in VLMs
Shuailong Wang, Xinyu Lyu, Shengming Yuan +3
Token-Pruning accelerates Vision-Language Models by removing redundant visual tokens, yet its safety implications remain underexplored. In this work, we present the first comprehen…
Janus-LoRA: A Balanced Low-Rank Adaptation for Continual Learning
Cheng Chen, Pengpeng Zeng, Yuyu Guo +3
Low-Rank Adaptation (LoRA) has emerged as a promising paradigm for Continual Learning. It independently updates its low-rank factors ( and ), creating a composite update to t…
Structure-aware Prompt Adaptation from Seen to Unseen for Open-Vocabulary Compositional Zero-Shot Learning
Yihang Duan, Jiong Wang, Pengpeng Zeng +5
The goal of Open-Vocabulary Compositional Zero-Shot Learning (OV-CZSL) is to recognize attribute-object compositions in the open-vocabulary setting, where compositions of both seen…
SkeleGuide: Explicit Skeleton Reasoning for Context-Aware Human-in-Place Image Synthesis
Chuqiao Wu, Jin Song, Yiyun Fei
Generating realistic and structurally plausible human images into existing scenes remains a significant challenge for current generative models, which often produce artifacts like…
TIMI: Training-Free Image-to-3D Multi-Instance Generation with Spatial Fidelity
Xiao Cai, Pengpeng Zeng, Ji Zhang +3
Precise spatial fidelity in Image-to-3D multi-instance generation is critical for downstream real-world applications. Recent work attempts to address this by fine-tuning pre-traine…
Benchmarking Few-shot Transferability of Pre-trained Models with Improved Evaluation Protocols
Xu Luo, Ji Zhang, Lianli Gao +2
Few-shot transfer has been revolutionized by stronger pre-trained models and improved adaptation algorithms.However, there lacks a unified, rigorous evaluation protocol that is bot…
Beyond the Majority: Long-tail Imitation Learning for Robotic Manipulation
Junhong Zhu, Ji Zhang, Jingkuan Song +2
While generalist robot policies hold significant promise for learning diverse manipulation skills through imitation, their performance is often hindered by the long-tail distributi…
Sim-and-Human Co-training for Data-Efficient and Scene-Generalizable Bimanual Manipulation
Kaipeng Fang, Weiqing Liang, Yuyang Li +5
Real-robot demonstrations are prohibitively expensive, while simulation data and real-world human demonstrations are both scalable but each leaves a distinct gap: simulation suffer…
From One-to-One to Many-to-Many: Dynamic Cross-Layer Injection for Deep Vision-Language Fusion
Cheng Chen, Yuyu Guo, Pengpeng Zeng +4
Vision-Language Models (VLMs) create a severe visual feature bottleneck by using a crude, asymmetric connection that links only the output of the vision encoder to the input of the…
MiVLA: Towards Generalizable Vision-Language-Action Model with Human-Robot Mutual Imitation Pre-training
Zhenhan Yin, Xuanhan Wang, Jiahao Jiang +8
While leveraging abundant human videos and simulated robot data poses a scalable solution to the scarcity of real-world robot data, the generalization capability of existing vision…
Pseudo-Label Refinement for Robust Wheat Head Segmentation via Two-Stage Hybrid Training
Jiahao Jiang, Zhangrui Yang, Xuanhan Wang +1
This extended abstract details our solution for the Global Wheat Full Semantic Segmentation Competition. We developed a systematic self-training framework. This framework combines…
Reversible Inversion for Training-Free Exemplar-guided Image Editing
Yuke Li, Lianli Gao, Ji Zhang +5
Exemplar-guided Image Editing (EIE) aims to modify a source image according to a visual reference. Existing approaches often require large-scale pre-training to learn relationships…
FlexAC: Towards Flexible Control of Associative Reasoning in Multimodal Large Language Models
Shengming Yuan, Xinyu Lyu, Shuailong Wang +3
Multimodal large language models (MLLMs) face an inherent trade-off between faithfulness and creativity, as different tasks require varying degrees of associative reasoning. Howeve…
A Survey on Efficient Vision-Language-Action Models
Zhaoshu Yu, Bo Wang, Pengpeng Zeng +7
Vision-Language-Action models (VLAs) represent a significant frontier in embodied intelligence, aiming to bridge digital knowledge with physical-world interaction. Despite their re…
An Empirical Analysis of VLM-based OOD Detection: Mechanisms, Advantages, and Sensitivity
Yuxiao Lee, Xiaofeng Cao, Wei Ye +3
Vision-Language Models (VLMs), such as CLIP, have demonstrated remarkable zero-shot out-of-distribution (OOD) detection capabilities, vital for reliable AI systems. Despite this pr…
SINDI: An Efficient Index for Sparse Vector Approximate Maximum Inner Product Search
Ruoxuan Li, Xiaoyao Zhong, Jiabao Jin +7
Sparse vector Maximum Inner Product Search (MIPS) is crucial in multi-path retrieval for Retrieval-Augmented Generation (RAG). Recent inverted index-based and graph-based algorithm…
Dynamic Pattern Alignment Learning for Pretraining Lightweight Human-Centric Vision Models
Xuanhan Wang, Huimin Deng, Ke Liu +3
Human-centric vision models (HVMs) have achieved remarkable generalization due to large-scale pretraining on massive person images. However, their dependence on large neural archit…
Shortcut Learning in Generalist Robot Policies: The Role of Dataset Diversity and Fragmentation
Youguang Xing, Xu Luo, Junlin Xie +3
Generalist robot policies trained on large-scale datasets such as Open X-Embodiment (OXE) demonstrate strong performance across a wide range of tasks. However, they often struggle…
Learning Generalizable and Efficient Image Watermarking via Hierarchical Two-Stage Optimization
Ke Liu, Xuanhan Wang, Qilong Zhang +2
Deep image watermarking, which refers to enabling imperceptible watermark embedding and reliable extraction in cover images, has been shown to be effective for copyright protection…
SafePTR: Token-Level Jailbreak Defense in Multimodal LLMs via Prune-then-Restore Mechanism
Beitao Chen, Xinyu Lyu, Lianli Gao +2
By incorporating visual inputs, Multimodal Large Language Models (MLLMs) extend LLMs to support visual reasoning. However, this integration also introduces new vulnerabilities, mak…
A Closer Look at Conditional Prompt Tuning for Vision-Language Models
Ji Zhang, Shihan Wu, Lianli Gao +3
Despite the great promise of Prompt Tuning (PT) in adapting large Vision-Language Pretrained Models (VLPMs) to downstream tasks, they often struggle to overcome the Base-New Tradeo…
EnhanceGraph: A Continuously Enhanced Graph-based Index for High-dimensional Approximate Nearest Neighbor Search
Xiaoyao Zhong, Jiabao Jin, Peng Cheng +5
Recently, Approximate Nearest Neighbor Search in high-dimensional vector spaces has garnered considerable attention due to the rapid advancement of deep learning techniques. We obs…
Reliable Few-shot Learning under Dual Noises
Ji Zhang, Jingkuan Song, Lianli Gao +2
Recent advances in model pre-training give rise to task adaptation-based few-shot learning (FSL), where the goal is to adapt a pre-trained task-agnostic model for capturing task-sp…
OmniCharacter: Towards Immersive Role-Playing Agents with Seamless Speech-Language Personality Interaction
Haonan Zhang, Run Luo, Xiong Liu +10
Role-Playing Agents (RPAs), benefiting from large language models, is an emerging interactive AI system that simulates roles or characters with diverse personalities. However, exis…