2 Papers Accepted at ECCV 2026

Held on 8 – 12 September in Malmö, Sweden, the 19th European Conference on Computer Vision (ECCV 2026) is a biennial premier research conference in Computer Vision and Machine Learning, recognised as one of the top-tier global venues for advancing the science and application of visual perception systems.

Congratulations to the following researchers from A*STAR Centre for Frontier AI Research (A*STAR CFAR) on having their papers accepted at ECCV 2026:

• Dr Nancy F. Chen, Group Lead (Ethical & Trust AI)
• Dr Zhang Jie, Innovation Lead

List of accepted papers:

1. Explicit Logic Channel for Validation and Enhancement of MLLMs on Zero-Shot Tasks
   Mei Chee Leong, Ying Gu, Hui Li Tan, Liyuan Li, Nancy Chen

   Advanced visual-language AI models often act as closed black boxes, making it hard to validate their performance on new tasks. To solve this, researchers introduced a parallel "Explicit Logic Channel" that mimics human reasoning using separate language and vision tools. This framework successfully improves model trustworthiness, validation, and selection without needing ground-truth data.

2. GlassGS: Geometry and Concept-Aware 3D Gaussian Splatting for Reflective Enclosures
   Mingxuan Cui, Yunrui Zhu, Wuqi Wang, Di Lin, Jianhua Zhang, Jie Zhang, Ming-Ming Cheng, Shengyong Chen, Qing Guo**
   
   This paper introduces GlassGS, a novel approach for 3D Gaussian Splatting reconstruction of objects behind glass. By identifying reflection-induced geometric inconsistencies and learning reflection-robust object representations, GlassGS enables more accurate and reliable 3D reconstruction in reflective environments.

More on ECCV 2026.