M2KR-Challenge Dataset

A multimodal retrieval dataset for image-to-document and image+text-to-document matching tasks.

Dataset Overview

This dataset contains two main subsets designed for multimodal retrieval challenges:

• challenge_data: Query data with images and optional text questions (6.42k samples)
• challenge_passage: Document collection with textual passages and associated web screenshot path (47.3k passages)

Dataset Structure

challenge_data (6,420 rows)

Columns:

• img_path: Image filename (string)
• instruction: Task instruction for description generation
• question: Optional text query (53% populated, if exist then it is a image+text-to-document retrieval task, else, it is a image-to-document task )
• question_id: Unique identifier (string)
• pos_item_ids: Sequence of positive item IDs, the ground truth passage_id (empty, removed for private test set)
• pos_item_contents: Sequence of positive item contents (empty, removed for private test set)

Task Types:

1. Image-to-Document Retrieval: When question is empty (image query)
2. Multimodal Retrieval: When question contains text (image + text query)

challenge_passage (47,300 rows)

Columns:

• passage_id: Unique passage identifier (string)
• passage_content: Textual description containing:
  • Image description
  • Structured details about persons (birth/death dates, occupations, locations, etc.)
• page_screenshot: Associated image filename (string)

For the retrieval task, you will need to retrieve the corresponding passage from the 47K-passage pool for each sample in challenge_data.

Images

The image data is provided in separate archives:

• Web_Image.zip.001
• Web_Image.zip.002
• Web_Image.zip.003

These archives contain the web screenshots corresponding to the document passages.

• query_images.zip
  Contains the query images used in the challenge.

---

References

Paper or resources for more information:

• Paper: https://arxiv.org/abs/2402.08327
• Project Page: https://preflmr.github.io/
• Huggingface Implementation: https://github.com/LinWeizheDragon/FLMR

Citation If our work helped your research, please kindly cite our paper for PreFLMR.

       
@inproceedings{lin-etal-2024-preflmr,
    title = "{P}re{FLMR}: Scaling Up Fine-Grained Late-Interaction Multi-modal Retrievers",
    author = "Lin, Weizhe  and
      Mei, Jingbiao  and
      Chen, Jinghong  and
      Byrne, Bill",
    editor = "Ku, Lun-Wei  and
      Martins, Andre  and
      Srikumar, Vivek",
    booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
    month = aug,
    year = "2024",
    address = "Bangkok, Thailand",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2024.acl-long.289",
    pages = "5294--5316",
    abstract = "Large Multimodal Models (LMMs) excel in natural language and visual understanding but are challenged by exacting tasks such as Knowledge-based Visual Question Answering (KB-VQA) which involve the retrieval of relevant information from document collections to use in shaping answers to questions. We present an extensive training and evaluation framework, M2KR, for KB-VQA. M2KR contains a collection of vision and language tasks which we have incorporated into a single suite of benchmark tasks for training and evaluating general-purpose multi-modal retrievers. We use M2KR to develop PreFLMR, a pre-trained version of the recently developed Fine-grained Late-interaction Multi-modal Retriever (FLMR) approach to KB-VQA, and we report new state-of-the-art results across a range of tasks. We also present investigations into the scaling behaviors of PreFLMR intended to be useful in future developments in general-purpose multi-modal retrievers.",
}