UniMUG 2027

Deadlines
Computer Vision/CORE Unranked

UniMUG 2027

First Workshop on Visual Storytelling with Unified Multimodal Understanding and Generation

1799046000000Orlando, FL, USAOfficial workshop site Site reachable

UniMUG 2027 is the first workshop on Visual Storytelling with Unified Multimodal Understanding and Generation, held in conjunction with WACV 2027. It aims to bridge the gap between video understanding and generation by exploring shared latent representations and cross-leveraging capabilities to advance both domains through joint research and evaluation.

Official CFP Back to deadlines Verified September 8, 2026

Paper fit

Contribution paths

A strong submission should clearly identify its contribution and evaluate it appropriately.

Full papers

Submissions must follow WACV conference style, with a maximum of 14 pages (excluding references).

Research areas in scope

01

Joint Understanding & Generation

Novel architectures for unified multimodal understanding and generationShared latent representations across understanding and generation tasksMulti-task learning frameworks bridging perception and synthesisMethods to train unified models for joint understanding and generation
02

Generation-Enhanced Understanding

Leveraging generative priors for improved video comprehensionWorld models for video reasoning and predictionGenerative pre-training for downstream understanding tasks
03

Understanding-Enhanced Generation

Incorporating semantic understanding signals into video generationControllable generation through scene understandingConsistency enforcement via understanding modules
04

Efficiency & Scalability

Teacher-student knowledge distillation from expert modelsEfficient unified architectures for resource-constrained deploymentModel compression techniques for joint understanding-generation models
05

Evaluation & Benchmarks

Novel metrics for unified model assessmentBenchmark datasets for joint understanding and generationHuman evaluation methodologies
06

Applications

Long-form video storytelling and narrative generationInteractive video content creationVideo summarization and recap generationMulti-modal generation and understandingCross-modal alignment

Policies worth checking twice

  • Double-blind review
  • Supplementary Materials: Optional additional materials permitted
  • Dual Submission Policy: To be specified based on WACV 2027 guidelines

Official sources

Compiled from the official call for papers. The organizers’ pages remain authoritative.

Last verified September 8, 2026