M-2KR
Emerging5papers using it
2024first seen
The 'M2KR' dataset is used to evaluate open-domain retrieval of Wikipedia-style passages from an image or image-plus-text query within the context of the Multimodal Document Retrieval Challenge.
Papers using M-2KR (5)
- Overview of the EReL@MIR 2025 Multimodal Document Retrieval Challenge (Track 1)Recurrence Meets Transformers For Universal Multimodal RetrievalWikiSeeker: Rethinking the Role of Vision-Language Models in Knowledge-Based Visual Question AnsweringA Multi-Granularity Retrieval Framework for Visually-Rich DocumentsPreFLMR: Scaling Up Fine-Grained Late-Interaction Multi-modal Retrievers