Leveraging Multi-View Image Sets for Unsupervised Intrinsic Image Decomposition and Highlight Separation

Abstract

We present an unsupervised approach for factorizing object appearance into highlight, shading, and albedo layers, trained by multi-view real images. To do so, we construct a multi-view dataset by collecting numerous customer product photos online, which exhibit large illumination variations that make them suitable for training of reflectance separation and can facilitate object-level decomposition. The main contribution of our approach is a proposed image representation based on local color distributions that allows training to be insensitive to the local misalignments of multi-view images. In addition, we present a new guidance cue for unsupervised training that exploits synergy between highlight separation and intrinsic image decomposition. Over a broad range of objects, our technique is shown to yield state-of-the-art results for both of these tasks.

Cite

Text

Yi et al. "Leveraging Multi-View Image Sets for Unsupervised Intrinsic Image Decomposition and Highlight Separation." AAAI Conference on Artificial Intelligence, 2020. doi:10.1609/AAAI.V34I07.6961

Markdown

[Yi et al. "Leveraging Multi-View Image Sets for Unsupervised Intrinsic Image Decomposition and Highlight Separation." AAAI Conference on Artificial Intelligence, 2020.](https://mlanthology.org/aaai/2020/yi2020aaai-leveraging/) doi:10.1609/AAAI.V34I07.6961

BibTeX

@inproceedings{yi2020aaai-leveraging,
  title     = {{Leveraging Multi-View Image Sets for Unsupervised Intrinsic Image Decomposition and Highlight Separation}},
  author    = {Yi, Renjiao and Tan, Ping and Lin, Stephen},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2020},
  pages     = {12685-12692},
  doi       = {10.1609/AAAI.V34I07.6961},
  url       = {https://mlanthology.org/aaai/2020/yi2020aaai-leveraging/}
}