Full long papers
8 to 14 pages (excluding references or supplementary materials), submitted via OpenReview, adhering to ECCV 2026 guidelines, and subject to double-blind review.
Computer Vision/CORE Unranked
11thABAW2026 2026
11th Workshop and Competition on Affective & Behavior Analysis in-the-wild
The 11th ABAW Workshop and Competition at ECCV 2026 is a premier forum for advancing multimodal analysis of human affect and behavior in unconstrained, real-world environments. It brings together researchers from computer vision, machine learning, psychology, robotics, and healthcare to present cutting-edge methods in facial and body behavior analysis, with a focus on multimodal fusion, temporal reasoning, and responsible AI.
Paper fit
Contribution paths
A strong submission should clearly identify its contribution and evaluate it appropriately.
Research areas in scope
01
Affective and Behavioral Analysis
facial expression (basic, compound or other) or micro-expression analysisfacial action unit detectionvalence-arousal estimationphysiological-based (e.g., EEG, EDA) affect analysisface recognition, detection or trackingbody recognition, detection or trackinggesture recognition or detectionpose estimation or trackingactivity recognition or trackinglip reading and voice understandingface and body characterization (e.g., behavioral understanding)characteristic analysis (e.g., gait, age, gender, ethnicity recognition)group understanding via social cues (e.g., kinship, non-blood relationships, personality)video, action and event understanding
02
Applications and Systems
digital human modelingbehaviour-aware, affective and social roboticshuman-robot interaction, collaboration and communicationrobot perception of human affect, behaviour, intention, attention and social signalsembodied AI agents, assistive robots and socially interactive robotsviolence detectionautonomous driving
03
Methodological and Ethical Challenges
domain adaptation, domain generalisation, few- or zero-shot learning for the above casesfairness, explainability, interpretability, trustworthiness, safety, privacy-awareness, bias mitigation and/or subgroup distribution shift analysis for the above casesediting, manipulation, image-to-image translation, style mixing, interpolation, inversion and semantic diffusion for all afore mentioned cases
Policies worth checking twice
- All submissions must be anonymous and conform to ECCV 2026 standards for double-blind review.
- Papers must be at least 8 pages long to be considered for publication.
- Participants in the MTL Challenge must register via email with official institutional addresses; team leads cannot be students.
- Participants in the AH Challenge must register via a form signed by a full-time faculty member (not a student).
- Teams must submit a link to a GitHub repository with their solution/source code, a 2-8 page pre-print (e.g., arXiv), and test set predictions.
- Top-performing teams must submit papers describing their approach for inclusion in ECCV 2026 proceedings.
- Use of Aff-Wild2 (A/V) database is prohibited in the MTL Challenge; only s-Aff-Wild2 and other public/private datasets (excluding Aff-Wild2) are allowed.
- Teams may use any pre-trained models except those pre-trained on Aff-Wild2.
Official sources
Compiled from the official call for papers. The organizers’ pages remain authoritative.
Last verified September 9, 2026