1 dataset found

Tags: Scene text visual question answering

Filter Results

  • ST-VQA

    ST-VQA dataset consists of 23,038 images with 31,791 question-answer pairs.
You can also access this registry using the API (see API Docs).