The 2nd EReL@MIR Workshop on Efficient Representation Learning for Multimodal Information Retrieval

Researchers proposed the 2nd EReL@MIR workshop for MM 2026 to advance efficient multimodal representation learning for information retrieval.

The workshop will focus on lowering the training, deployment, and inference costs of multimodal models used in retrieval applications.

It aims to bring together work on practical efficiency methods, evaluation metrics, and benchmarks for cross-modal retrieval, web search, and recommender systems.

The effort responds to the high effectiveness but heavy cost of adapting large multimodal foundation models such as CLIP, LLaVA, and Qwen.