Official project page of TransMatting (ECCV2022)
The extended paper TransMatting: Tri-token Equipped Transformer Model for Image Matting is published on arXiv. We further propose TransMattingV2, which extends tri-token to both convolutional networks and Transformer. Various experiments demonstrate that it could boost both Transformer and convolutional networks, indicating that the proposed tri-token has the potential to be a new paradigm for image matting.
Quantitative Results on Composition-1k
| Model | SAD | MSE | Grad | Conn | Results |
|---|---|---|---|---|---|
| TransMattingV1 | 24.96 | 4.58 | 9.72 | 20.16 | GoogleDrive |
Our dataset Transparent-460 is now open to the public. If you want to use it for research rather than commercial activity, you can directly download it using the following link: Transparent-460
Please note that you are still required to send an email to Huanqia Cai ([email protected]) to inform us of your usage. Your email MUST be sent from a valid university or research institute account and MUST include the following text:
Subject: Application to download the Transparent-460 Database
Name: <Your first and last name>
Affiliation: <University or institute where you work>
Department: <Your department>
Position: <Your job title>
Email: <Must be the email at the above mentioned institution>
I have read and agree to the terms and conditions specified in the Transparent-460 database webpage.
This database will only be used for research purposes.
I will not make any part of this database available to a third party.
I'll not sell any part of this database or make any profit from its use.
The Transparent-460 database is available for non-commercial research purposes only.
-
All images of the Transparent-460 database are obtained from the Internet and are not the authors' property. The authors are not responsible for these images' content or meaning.
-
You agree not to reproduce, duplicate, copy, sell, trade, resell or exploit any portion of the images and any portion of derived data for any commercial purposes.
-
You agree not to further copy, publish or distribute any portion of the Transparent-460 database. Except for internal use at a single site within the same organization, it is allowed to make copies of the dataset.
-
The authors reserve the right to terminate your access to the Transparent-460 database at any time.
@inproceedings{cai2022TransMatting,
title={TransMatting: Enhancing Transparent Objects Matting with Transformers},
author={Cai, Huanqia and Xue, Fanglei, and Xu, Lele and Guo, Lili},
booktitle={European Conference on Computer Vision (ECCV)},
year={2022},
}