Venue and Schedule

Workshop schedule

Tuesday 8 September 2026 · 8:30 am – 12:30 pm

  • 8:30 amWelcome and opening remarks10 min
  • 8:40 amKeynote 1 — Elisa Ricci40 min
  • 9:20 amKeynote 2 — Steffen Schneider40 min
  • 10:00 amCoffee break30 min
  • 10:30 amPoster session75 min
  • 11:45 amKeynote 3 — Andrew King40 min
  • 12:25 pmClosing remarks5 min
Keynote Posters Chair Break

Accepted regular papers

5

  1. A Compositional Analysis of Cross-Lingual Math Reasoning in Efficient Vision-Language Models

    Kerem Tuzel1, Ozan Bayiz2

    1Fremont Union High School District  ·  2University of California, Berkeley

  2. A Synonym Away from Failure: Measuring and Mitigating Word-Choice Sensitivity in Promptable Concept Segmentation

    Stefano Samele, Michele Cavicchioli, Nico Catalano, Matteo Matteucci

    Politecnico di Milano

  3. Debiasing Text-to-Image Evaluation via Implicit Cultural Alignment Reward Modeling

    Boan Chang1, Yu-Chih Chen2

    1National Tsing Hua University  ·  2National Yang Ming Chiao Tung University

  4. Fine-Grained Bias Manipulation of Text-to-Image Models via Counterfactual Data Synthesis

    Takuto Onikubo1, Xin Zhang2, Joey Tianyi Zhou3, Shin'ichi Satoh4

    1The University of Tokyo  ·  2A*STAR  ·  3Singapore Management University; A*STAR Centre for Frontier AI Research  ·  4National Institute of Informatics

  5. Localizing to Debias: A Patch-Level Benchmark and Baseline for Weakly Supervised Spatial Anomaly Detection

    Sara Abdulaziz, Abdulrahman Al-Abri, Giacomo D'Amicantonio, Egor Bondarev

    Eindhoven University of Technology

Accepted extended abstracts non-archival

6

  1. CFM: Language-aligned Concept Foundation Model for Vision

    Kai Wittenmayer, Sukrut Rao, Amin Parchami-Araghi, Bernt Schiele

    Max Planck Institute for Informatics, Saarland Informatics Campus

  2. Do Vision-Language Models See or Remember? Probability-Level Evidence of Prior Bias on Counterfactual Images

    Rabia Abu Sufian, Ayat Khudoir, Stella Graßhof, Luca Rossi

    IT University of Copenhagen

  3. Feature Alignment for Scalable B-cosification of Foundational Vision Transformers

    Raphael Maser1,2, Siddhartha Gairola1, Sukrut Rao1,2, Bernt Schiele1

    1Max Planck Institute for Informatics, Saarland Informatics Campus  ·  2Universität des Saarlandes

  4. Open Set Domain Adaptation Under Background Distribution Shift

    Shravan S. Chaudhari1, Yoav Wald2, Suchi Saria1

    1Johns Hopkins University  ·  2Technion – Israel Institute of Technology; New York University

  5. Spectral Analysis of Intersectional Bias in Diffusion Models

    Silpa Vadakkeeveetil Sreelatha1, Anisha Saha2, Muhammad Awais1, Anjan Dutta1, Serge Belongie3, Stella Frank4, Vésteinn Snæbjarnarson3

    1University of Surrey  ·  2Max Planck Institute for Informatics, Saarland Informatics Campus  ·  3University of Copenhagen  ·  4Technical University of Denmark

  6. Who Defines Fairness? Target-Based Prompting for Auditable Debiasing in Text-to-Image Generation

    Marzia Binta Nizam, James Davis

    University of California, Santa Cruz