https://github.com/dddb11/MVSS-Net
Un-official inplementation of MVSS-Net which was proposed in ICCV 2021 by Chengbo Dong et al, includes code for training!.
What need to point out is that, the original repo is laking of the training code, links is here: OFFICAL MVSS-Net link
We need to point out that many researchers reported that they faild to reproduce the result in their paper. This is also our goal when building this repo, but in the end, our re-implementation also fails and the metrics are far from the metrics in the paper on the Columbia and COVER datasets. You could disscuss this phenomenon with researchers through this links: Zhihu(a Chinese forum); Issues of the offical repo
Still on Working...
Still on Working...
Still on Working...
If you want to train this Model with the CASIAv2 dataset, we provide a revised version of CASIAv2 datasets, which corrected several mistakes in the original datasets provided by the author. Details can find in the link shown below:
[1] Chen, X., Dong, C., Ji, J., Cao, J., & Li, X. (2021). Image Manipulation Detection by Multi-View Multi-Scale Supervision. 2021 IEEE/CVF International Conference on Computer Vision (ICCV), 14165–14173. https://doi.org/10.1109/ICCV48922.2021.01392