[Apologies for cross-posting] Dear colleagues, There is still plenty of time to participate in the ChaLearn UDIVA-HHOI Challenge @ ECCV 2026, organized within the CONTEXTUS Workshop at ECCV 2026. Submission deadline: July 12th, 2026 The challenge focuses on advancing research in context-aware human behavior understanding and socially grounded multimodal intelligence, encouraging methods that go beyond isolated action recognition toward understanding how people collaborate, coordinate, anticipate, and influence one another in real-world environments. Participants will work with the newly released UDIVA-HHOI dataset, a rich multimodal corpus featuring: * non-scripted dyadic collaborative interactions, * synchronized audio, video, and transcripts, * one exocentric and two egocentric views, * contextual metadata and social interaction cues, * procedural task information, * annotations of verbal and non-verbal human-human-object interaction events, * goals, intentions, and causal relationships. The challenge includes 5 tracks: * Track 1: Multimodal Exocentric Event Recognition * Track 2: Multimodal Egocentric Event Recognition * Track 3: Multimodal Exocentric Event Anticipation * Track 4: Multimodal Egocentric Event Anticipation * Track 5: Multimodal Exocentric Causal Event Grounding The top-ranked team in each track will receive a $1,000 USD award, along with certificates and the opportunity to present and publish their work at the ECCV 2026 CONTEXTUS Workshop. Data, starting kit, track descriptions, and metrics are already available, and baseline models will be released soon. We strongly encourage participation from researchers and students working on: * multimodal learning, * video understanding, * egocentric vision, * embodied AI, * social signal processing, * event anticipation, * causal reasoning, * human behavior understanding. We look forward to seeing new advances toward AI systems capable of understanding human interactions, intentions, and collaborative behavior in complex real-world scenarios. Challenge website: https://lap.chalearn.eu/public/udiva-hhoi-challenge-eccv26 Workshop website: https://lap.chalearn.eu/public/ECCV26-CONTEXTUS Supported by ChaLearn, SurfingTech, and Google. Organizers: Cristina Palmero (King's College London) Sergio Escalera (Universitat de Barcelona & Computer Vision Center) Albert Clapés (Universitat de Barcelona) Xavier Baró (Universitat de Barcelona) Daniele Berardini (Istituto Italiano di Tecnologia) Hugo Jair Escalante (The University of Texas at El Paso & INAOE) Vittorio Murino (Istituto Italiano di Tecnologia & University of Verona) Challenge chairs: Jeanfed Ramírez Lima, Luis J. Arellano Advisory board: Isabelle Guyon, Jeffrey Cohn