English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Inverse rendering is the problem of decomposing an image into its intrinsiccomponents, i.e. albedo, normal and lighting. To solve this ill-posed problemfrom single image, state-of-the-art methods in shape from shading mostly resortto supervised training on all the components on either synthetic or realdatasets. Here, we propose a new self-supervised training paradigm that 1)reduces the need for full supervision on the decomposition task and 2) takesinto account the relighting task. We introduce new self-supervised loss termsthat leverage the consistencies between multi-lit images (images of the samescene under different illuminations). Our approach is applicable to multi-litdatasets. We apply our training approach in two settings: 1) train on a mixtureof synthetic and real data, 2) train on real datasets with limited supervision.We show-case the effectiveness of our training paradigm on both intrinsicdecomposition and relighting and demonstrate how the model struggles in bothtasks without the self-supervised loss terms in limited supervision settings.We provide results of comprehensive experiments on SfSNet, CelebA and Photofacedatasets and verify the performance of our approach on images in the wild.

Joint Learning of Portrait Intrinsic Decomposition and Relighting