VLM2Vec-V2: Advancing Multimodal Embedding for Videos, Images, and Visual Documents

TL;DR

VLM2Vec-V2 introduces a unified multimodal embedding framework supporting diverse visual inputs, including videos and documents, with a new benchmark and improved performance across multiple retrieval and understanding tasks.

Contribution

The paper presents VLM2Vec-V2, a novel model and benchmark extending multimodal embeddings to videos and visual documents, enhancing applicability and performance.

Findings

VLM2Vec-V2 outperforms prior models on new video and document retrieval tasks.

The benchmark MMEB-V2 covers five new multimodal tasks.

The model improves existing image benchmarks while supporting diverse visual modalities.

Abstract

Multimodal embedding models have been crucial in enabling various downstream tasks such as semantic similarity, information retrieval, and clustering over different modalities. However, existing multimodal embeddings like VLM2Vec, E5-V, GME are predominantly focused on natural images, with limited support for other visual forms such as videos and visual documents. This restricts their applicability in real-world scenarios, including AI agents, multi-modal search and recommendation, and retrieval-augmented generation (RAG). To close this gap, we propose VLM2Vec-V2, a unified framework for learning embeddings across diverse visual forms. First, we introduce MMEB-V2, a comprehensive benchmark that extends MMEB with five new task types: visual document retrieval, video retrieval, temporal grounding, video classification and video question answering - spanning text, image, video, and visual document inputs. Next, we train VLM2Vec-V2, a general-purpose embedding model that supports text, image, video, and visual document inputs. Extensive experiments show that VLM2Vec-V2 achieves strong performance not only on the newly introduced video and document retrieval tasks, but also improves over prior baselines on the original image benchmarks. Through extensive evaluation, our study offers insights into the generalizability of various multimodal embedding models and highlights effective strategies for unified embedding learning, laying the groundwork for more scalable and adaptable representation learning in both research and real-world settings.