CoCA Dataset

📦 Common Category Aggregation (CoCA) dataset

CoCA dataset consists of 80 categories with 1295 images, covering everyday indoor and outdoor scenes. It is worth noting that these categories are outright staggered with Microsoft COCO. In our CoCA dataset, except for the co-salient object(s), each image contains at least one extraneous salient object, which enables the dataset to better evaluate the models’ ability of discovering co-salient object(s) among multiple foregrounds. We provide four levels of annotations: class level, bounding box level, object level, and instance level. Different levels of annotations of our dataset corresponds to different tasks, such as co-localization, few-shot object segmentation, co-saliency detection, and instance co-segmentation.

Cinque Terre Cinque Terre

⏬ Download

Download our CoCA dataset at Baidu Pan (iqf3), and Google Drive. Our dataset is freely available for research, but not for commercial use.

🗂 Co-Saliency Maps

We are collecting perdicted maps to facilitate follow-up research. These methods provide predictions on CoCA, CoSOD3k, and CoSal2015. We encourage researchers to add your method to this list. You just need to send your paper link, publisher, method abbreviation, and predicted probability maps on the three datatsets to my email.

Year Publisher Methods Baidu Pan Google Drive
2013 TIP CBCD Key: 158e Google drive link
2019 CVPR CSMG Key: 42rj Google drive link
2020 CVPR GCAGC Key: ij29 Google drive link
2020 ECCV GICD Key: puji Google drive link
2020 arXiv CoEG-Net Key:7ocj Google drive link

📐 Evaluation Tool

We released a GPU-accelerated evaluation tool (based on Python & PyTorch) for co-saliency detection task. It can automatically evaluate 8 metrics and draw 4 types of curves.

🤖️ Gradient-Induced Co-Saliency Detection (GICD)

📃 Citation

        title={Gradient-Induced Co-Saliency Detection},
        author={Zhang, Zhao and Jin, Wenda and Xu, Jun and Cheng, Ming-Ming},
        booktitle={European Conference on Computer Vision (ECCV)},