13 papers
Readout Blindness: VLM Scores Miss the Spatial Direction Their Frozen Encoders Retain
Guangyuan Li, Tianming Du, Yan Jiang +2
CLIP-like vision-language models remain a cornerstone of multimodal systems, yet their scores stay near chance on directed spatial relations, such as whether one object is left of…
GameDirector: Decoupling Gameplay Logic from Rendering for Player-Configurable Game World Models
Zijun Lin, Zhiyang Deng, Yuzhe Wu +2
Recent game world models support realistic visual simulation and interactive gameplay based on player inputs. However, they typically learn environment dynamics from pixel-level su…
Video-OPSD: Exploiting Privileged Visual Evidence for On-Policy Self-Distillation in Video Large Language Models
Ziyue Wang, Shiqi Huang, Weiwen Xu +2
On-policy self-distillation (OPSD) has recently emerged as an effective post-training paradigm that improves policy optimization through dense token-level supervision from a privil…
StatePlay: State-Aware Game World Models for Mechanics-Consistent Generation
Zijun Lin, Zeqing Wang, Cheston Tan +2
Recent game world models can generate visually realistic and interactive environments conditioned on player actions. However, games are not defined by pixels alone; they are govern…
EvoVid: Temporal-Centric Self-Evolution for Video Large Language Models
Shiqi Huang, Ziyue Wang, Zhongrong Zuo +3
Recent Video Large Language Models (Video-LLMs) have demonstrated strong capabilities in video reasoning through reinforcement learning (RL). However, existing RL pipelines rely he…
GEditBench v2: A Human-Aligned Benchmark for General Image Editing
Zhangqi Jiang, Zheng Sun, Xianfang Zeng +7
Recent advances in image editing have enabled models to handle complex instructions with impressive realism. However, existing evaluation frameworks lag behind: current benchmarks…
RSGround-R1: Rethinking Remote Sensing Visual Grounding through Spatial Reasoning
Shiqi Huang, Shuting He, Bihan Wen
Remote Sensing Visual Grounding (RSVG) aims to localize target objects in large-scale aerial imagery based on natural language descriptions. Owing to the vast spatial scale and hig…
FailSafe: Reasoning and Recovery from Failures in Vision-Language-Action Models
Zijun Lin, Jiafei Duan, Haoquan Fang +4
Recent advances in robotic manipulation have integrated low-level robotic control into Vision-Language Models (VLMs), extending them into Vision-Language-Action (VLA) models. Altho…
The 1st International Workshop on Disentangled Representation Learning for Controllable Generation (DRL4Real): Methods and Results
Qiuyu Chen, Xin Jin, Yue Song +45
This paper reviews the 1st International Workshop on Disentangled Representation Learning for Controllable Generation (DRL4Real), held in conjunction with ICCV 2025. The workshop a…
SAVER: Mitigating Hallucinations in Large Vision-Language Models via Style-Aware Visual Early Revision
Zhaoxu Li, Chenqi Kong, Yi Yu +6
Large Vision-Language Models (LVLMs) recently achieve significant breakthroughs in understanding complex visual-textual contexts. However, hallucination issues still limit their re…
SCORE: Scene Context Matters in Open-Vocabulary Remote Sensing Instance Segmentation
Shiqi Huang, Shuting He, Huaiyuan Qin +1
Most existing remote sensing instance segmentation approaches are designed for close-vocabulary prediction, limiting their ability to recognize novel categories or generalize acros…
GroundFlow: A Plug-in Module for Temporal Reasoning on 3D Point Cloud Sequential Grounding
Zijun Lin, Shuting He, Cheston Tan +1
Sequential grounding in 3D point clouds (SG3D) refers to locating sequences of objects by following text instructions for a daily activity with detailed steps. Current 3D visual gr…
ZoRI: Towards Discriminative Zero-Shot Remote Sensing Instance Segmentation
Shiqi Huang, Shuting He, Bihan Wen
Instance segmentation algorithms in remote sensing are typically based on conventional methods, limiting their application to seen scenarios and closed-set predictions. In this wor…