dugMatting: Decomposed-Uncertainty-Guided Matting

Jiawei Wu, Changqing Zhang, Zuoyong Li, Huazhu Fu, Xi Peng, Joey Tianyi Zhou
Proceedings of the 40th International Conference on Machine Learning, PMLR 202:37846-37859, 2023.

Abstract

Cutting out an object and estimating its opacity mask, known as image matting, is a key task in image and video editing. Due to the highly ill-posed issue, additional inputs, typically user-defined trimaps or scribbles, are usually needed to reduce the uncertainty. Although effective, it is either time consuming or only suitable for experienced users who know where to place the strokes. In this work, we propose a decomposed-uncertainty-guided matting (dugMatting) algorithm, which explores the explicitly decomposed uncertainties to efficiently and effectively improve the results. Basing on the characteristic of these uncertainties, the epistemic uncertainty is reduced in the process of guiding interaction (which introduces prior knowledge), while the aleatoric uncertainty is reduced in modeling data distribution (which introduces statistics for both data and possible noise). The proposed matting framework relieves the requirement for users to determine the interaction areas by using simple and efficient labeling. Extensively quantitative and qualitative results validate that the proposed method significantly improves the original matting algorithms in terms of both efficiency and efficacy.

Cite this Paper

BibTeX
@InProceedings{pmlr-v202-wu23y, title = {dug{M}atting: Decomposed-Uncertainty-Guided Matting}, author = {Wu, Jiawei and Zhang, Changqing and Li, Zuoyong and Fu, Huazhu and Peng, Xi and Zhou, Joey Tianyi}, booktitle = {Proceedings of the 40th International Conference on Machine Learning}, pages = {37846--37859}, year = {2023}, editor = {Krause, Andreas and Brunskill, Emma and Cho, Kyunghyun and Engelhardt, Barbara and Sabato, Sivan and Scarlett, Jonathan}, volume = {202}, series = {Proceedings of Machine Learning Research}, month = {23--29 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v202/wu23y/wu23y.pdf}, url = {https://proceedings.mlr.press/v202/wu23y.html}, abstract = {Cutting out an object and estimating its opacity mask, known as image matting, is a key task in image and video editing. Due to the highly ill-posed issue, additional inputs, typically user-defined trimaps or scribbles, are usually needed to reduce the uncertainty. Although effective, it is either time consuming or only suitable for experienced users who know where to place the strokes. In this work, we propose a decomposed-uncertainty-guided matting (dugMatting) algorithm, which explores the explicitly decomposed uncertainties to efficiently and effectively improve the results. Basing on the characteristic of these uncertainties, the epistemic uncertainty is reduced in the process of guiding interaction (which introduces prior knowledge), while the aleatoric uncertainty is reduced in modeling data distribution (which introduces statistics for both data and possible noise). The proposed matting framework relieves the requirement for users to determine the interaction areas by using simple and efficient labeling. Extensively quantitative and qualitative results validate that the proposed method significantly improves the original matting algorithms in terms of both efficiency and efficacy.} }
Endnote
%0 Conference Paper %T dugMatting: Decomposed-Uncertainty-Guided Matting %A Jiawei Wu %A Changqing Zhang %A Zuoyong Li %A Huazhu Fu %A Xi Peng %A Joey Tianyi Zhou %B Proceedings of the 40th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2023 %E Andreas Krause %E Emma Brunskill %E Kyunghyun Cho %E Barbara Engelhardt %E Sivan Sabato %E Jonathan Scarlett %F pmlr-v202-wu23y %I PMLR %P 37846--37859 %U https://proceedings.mlr.press/v202/wu23y.html %V 202 %X Cutting out an object and estimating its opacity mask, known as image matting, is a key task in image and video editing. Due to the highly ill-posed issue, additional inputs, typically user-defined trimaps or scribbles, are usually needed to reduce the uncertainty. Although effective, it is either time consuming or only suitable for experienced users who know where to place the strokes. In this work, we propose a decomposed-uncertainty-guided matting (dugMatting) algorithm, which explores the explicitly decomposed uncertainties to efficiently and effectively improve the results. Basing on the characteristic of these uncertainties, the epistemic uncertainty is reduced in the process of guiding interaction (which introduces prior knowledge), while the aleatoric uncertainty is reduced in modeling data distribution (which introduces statistics for both data and possible noise). The proposed matting framework relieves the requirement for users to determine the interaction areas by using simple and efficient labeling. Extensively quantitative and qualitative results validate that the proposed method significantly improves the original matting algorithms in terms of both efficiency and efficacy.
APA
Wu, J., Zhang, C., Li, Z., Fu, H., Peng, X. & Zhou, J.T.. (2023). dugMatting: Decomposed-Uncertainty-Guided Matting. Proceedings of the 40th International Conference on Machine Learning, in Proceedings of Machine Learning Research 202:37846-37859 Available from https://proceedings.mlr.press/v202/wu23y.html.

Related Material