GazeVQA
收藏数据集概述
数据集名称
Gaze-grounded Visual Question Answering Dataset (GazeVQA)
数据集介绍
GazeVQA 是由 Shun Inadumi, Seiya Kawano, Akishige Yuguchi, Yasutomo Kawanishi, Koichiro Yoshino 等人提出,旨在澄清日语中的模糊问题。该数据集在 LREC-COLING 2024 会议上发布。
数据集内容
GazeVQA 包含 17,276 个问题/答案对,这些数据来源于 Gazefollow 和 COCO 数据集。
数据集格式
-
QA格式: json [ { "image_id": COCO 图像识别码, "qa_id": QA 样本识别码, "question": 问题, "answer": 答案(测试集有十个答案), "c_question": 澄清问题(仅测试集有) }, ... ]
-
QA属性格式: json { "qa_id":{ "gf_path": Gazefollow 图像和注视点识别码, "bboxes": COCO 注视目标边界框注释, [ [x1, y1, w, h], # obj1 [x1, y1, w, h], # obj2 ... ], "objects": COCO 注视目标对象标签注释 [obj1, obj2, ...] }, ... }
许可证
本数据集遵循 Creative Commons Attribution 4.0 License。
引用信息
bibtex @inproceedings{inadumi-etal-2024-gaze-grounded, title = "A Gaze-grounded Visual Question Answering Dataset for Clarifying Ambiguous {J}apanese Questions, author = "Shun Inadumi and Seiya Kawano and Akishige Yuguchi and Yasutomo Kawanishi and Koichiro Yoshino", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", pages = "558--571" year = "2024" }




