Semi-Supervised Video Instance Segmentation

Challenges

Densely Annotated VIdeo Segmentation (DAVIS): CVPR 2017-2020

Results on the CVPR DAVIS Challenge

We achieved the 4th place on the DAVIS Challenge 2020: [Leader Board 2020] (out of 11 team submissions)

We achieved the 3rd place on the DAVIS Challenge 2019: [Leader Board 2019] (out of 15 team submissions)

We achieved the 6th place on the DAVIS Challenge 2018: [Leader Board 2018] (out of 41 team submissions)

We achieved the 3rd place on the DAVIS Challenge 2017: [Leader Board 2017] (out of 22 team submissions)

Demonstration

Wonderland Data

Examples

Downloads

Pre-Computed Results: [MRGIS-2020] [GIS-2019] [CIS-2018] [IRIF-2017]
Wonderland Data: [WD-DAVIS-2017-TrainVal]
Lucid Data: [Lucid-DAVIS-2017-Full]

If you use our Wonderland Data, please cite following publications:

Publications

Trung-Nghia Le, Tam V. Nguyen, Minh-Triet Tran, "Contextual Guided Segmentation Framework for Semi-supervised Video Instance Segmentation", Journal of Machine Vision and Applications (MVA), 2022. [PDF]
Minh-Triet Tran, Trung-Hieu Hoang, Tam V. Nguyen, Trung-Nghia Le, E-Ro Nguyen, Minh-Quan Le, Hoang-Phuc Nguyen-Dinh, Xuan-Nhat Hoang, Minh N. Do, "Multi-Referenced Guided Instance Segmentation Framework for Semi-supervised Video Instance Segmentation", CVPR Workshop on DAVIS Challenge on Video Object Segmentation, US, 2020. [PDF] [Presentation]
Minh-Triet Tran, Trung-Nghia Le, Tam V. Nguyen, Vinh Ton-That, Trung-Hieu Hoang, Ngoc-Minh Bui, Trong-Le Do, Quoc-An Luong, Vinh-Tiep Nguyen, Duc Anh Duong, Minh N. Do, "Guided Instance Segmentation Framework for Semi-Supervised Video Instance Segmentation", CVPR Workshop on DAVIS Challenge on Video Object Segmentation, US, 2019. [PDF] [Poster]
Minh-Triet Tran, Vinh Ton-That, Trung-Nghia Le, Khac-Tuan Nguyen, Tu V. Ninh, Tu-Khiem Le, Vinh-Tiep Nguyen, Tam V. Nguyen, Minh N. Do, “Context-based Instance Segmentation in Video Sequence,” CVPR Workshop on DAVIS Challenge on Video Object Segmentation, US, 2018. [PDF] [Poster]
Minh-Triet Tran, Tu V. Ninh, Tu-Khiem Le, Vinh Ton-That, Khac-Tuan Nguyen, Trung-Nghia Le, Tam V. Nguyen, "Instance Segmentation in Video with Human-Pose Guidance and Data Augmentation", Scientific Conference of University of Science, VNU-HCM, Vietnam, 2018 (In Vietnamese). [PDF]
Trung-Nghia Le, Khac-Tuan Nguyen, Manh-Hung Nguyen-Phan, That-Vinh Ton, Toan-Anh Nguyen, Xuan-Son Trinh, Quang-Hieu Dinh, Vinh-Tiep Nguyen, Anh-Duc Duong, Akihiro Sugimoto, Tam V. Nguyen, Minh-Triet Tran, “Instance Re-Identification Flow for Video Object Segmentation,” CVPR Workshop on DAVIS Challenge on Video Object Segmentation, US, 2017. [PDF] [Presentation]

License

Our data are published under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 License. You must not use this work for commercial purposes. If you alter or build upon this work, you have to distribute the resulting work only under the same license. If you are interested in commercial usage, please contact us.

Google Sites

Report abuse