ACM Multimedia 2026

ACM Multimedia 2026 — the 34th ACM International Conference on Multimedia — takes place from 10 to 14 November 2026 in Rio de Janeiro, Brazil. It is the flagship gathering of the multimedia research community and one of the most important venues for multimodal AI, drawing academic researchers, PhD students, and media-technology engineers who work across images, video, audio, speech, music, sensor and social data. Anyone publishing or building at the intersection of generative media and machine perception will find this the field's central annual meeting.

The 2026 programme

ACM Multimedia organises its work around the integration of information across modalities, and the 2026 edition follows the conference's recently updated scientific scope, structured into four major themes. The technical programme spans multimedia analysis and retrieval, multimodal fusion, vision-and-language, audio and music understanding, and the generative methods that now define so much of the field. Recurring strands include human-centred and affective computing, multimedia systems and streaming, 3D and the metaverse, and the trust questions surrounding deepfakes, content provenance and media authenticity.

Beyond the main paper track, ACM Multimedia is known for a distinctive set of components that give it real range:

• The Multimedia Grand Challenges, where industry and academic teams compete on concrete, applied problems posed by the community.
• The Brave New Ideas track, which deliberately rewards high-risk, high-reward research directions over incremental results.
• An Open Source Software Competition, tutorials, panels, a doctoral symposium, and an arts and demonstrations programme that connects researchers with practitioners.

Keynotes typically feature leading researchers from across multimedia, computer vision, and multimodal learning; the 2026 line-up is confirmed on the official site as the programme is finalised. A practical note for authors: from 1 January 2026 ACM has moved fully to Open Access, so publication terms and article-processing-charge waivers differ from prior years — details are published through ACM and the conference site.

Who should attend

ACM Multimedia is built for researchers and graduate students working on multimedia analysis, multimodal large language models, generative video and audio, retrieval, and cross-modal understanding. It is equally relevant to engineers from the media, streaming and creative-technology industries who need to track where the research frontier is heading, and to anyone working on synthetic-media detection, watermarking and content authenticity. The Grand Challenges and demonstrations make it a useful venue for applied teams, while the doctoral symposium and student programmes support early-career researchers entering the field.

Why it matters in 2026

Multimedia research is where some of the most visible AI of 2026 actually lives. Text-to-image and text-to-video generation, multimodal models that reason jointly over text, vision and sound, and the rapid maturing of audio and music generation all sit squarely within this community's remit. ACM Multimedia is where the methods behind generative video and cross-modal understanding are advanced, benchmarked and critiqued — not just demonstrated.

The conference also carries a serious responsibility strand. As synthetic media becomes cheap and convincing, the same researchers building generative systems are developing the technical countermeasures: deepfake detection, provenance signalling and watermarking, and standards for content authenticity. ACM Multimedia is one of the few venues where the generation and the defence sides of the field share a programme, which makes it a clear read on how the community is balancing capability against misuse.

Hosting the 2026 edition in Rio de Janeiro is significant in its own right. Bringing an ACM flagship to Brazil widens participation across Latin America's growing AI and multimedia research community, a region historically under-represented at top-tier venues. For a conference whose subject is how machines perceive and create across modalities, broadening who is in the room is part of the point.

A culture of benchmarks and reproducibility

Part of what makes ACM Multimedia influential is its emphasis on shared tasks, datasets and reproducible results. The Grand Challenges put concrete, industry-posed problems in front of the community each year, while the Open Source Software Competition and reproducibility tracks reward work others can build on rather than one-off demonstrations. Dozens of focused workshops and tutorials sit alongside the main programme, giving newcomers an accessible entry point into specialised areas such as multimodal retrieval, affective computing or media forensics. For practitioners, this culture means the methods presented here tend to arrive with code, data and benchmarks attached.

Attending and registration

ACM Multimedia 2026 is an in-person conference in Rio de Janeiro across five days, combining oral, video and poster presentations with tutorials, workshops and the satellite tracks. Submission deadlines for the main track, workshops and Grand Challenges run well ahead of the November dates, and registration and current pricing are handled through the official conference website at 2026.acmmm.org. Researchers planning to submit should watch the call-for-papers timelines, while those attending to follow the field can register once general registration opens.