Cops-Ref stands for Compositional Referring Expression Comprehension, which is a visual reasoning image dataset about the object reference understanding. The dataset contains 75,299 real images, 148,712 text descriptions, and 1,307,885 candidate regions. This dataset has two main features. One is a new text generation engine that can combine reasoning logic and visual features to generate text descriptions of varying degrees of complexity. The other is a new test setting that interferes with semantically similar visual images during the test.

Citation

@inroceedings{Chen_2020_CVPR, author = {Chen, Zhenfang and Wang, Peng and Ma, Lin and Wong, Kwan-Yee~K. and Wu, Qi}, title = {Cops-Ref: A New Dataset and Task on Compositional Referring Expression Comprehension}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month = {June}, year = {2020} }