We propose a self-supervised method for image relighting of single view images in the wild. The method is based on an auto-encoder which deconstructs an image into two separate encodings, relating to the scene illumination and content, respectively. In order to disentangle this embedding information without supervision, we exploit the assumption that some augmentation operations do not affect the image content and only affect the direction of the light. A novel loss function, called spherical harmonic loss, is introduced that forces the illumination embedding to convert to a spherical harmonic vector. We train our model on large-scale datasets such as Youtube 8M and CelebA. Our experiments show that our method can correctly estimate scene illumination and realistically re-light input images, without any supervision or a prior shape model. Compared to supervised methods, our approach has similar performance and avoids common lighting artifacts.
RelightVid: Temporal-Consistent Diffusion Model for Video Relighting
Shanghai AI Laboratory（上海人工智能实验室）；Fudan University（复旦大学）；Shanghai Jiao Tong University（上海交通大学）；Zhejiang University（浙江大学）；Stanford University（斯坦福大学）；The Chinese University of Hong Kong（香港中文大学）
MIT CSAIL（麻省理工学院计算机科学与人工智能实验室）；Adobe Research（奥多比研究院）
Collections of images under a single, uncontrolled illumination have enabled the rapid advancement of core computer vision tasks like classification, detection, and segmentation. But even with modern learning techniques, many inverse problems involving lighting and material understanding remain too severely ill-posed to be solved with single-illumination datasets. To fill this gap, we introduce a new multi-illumination dataset of more than 1000 real scenes, each captured under 25 lighting conditions. We demonstrate the richness of this dataset by training state-of-the-art models for three chal…
University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）；Apple Inc.（苹果公司）
Inverse rendering is the problem of decomposing an image into its intrinsic components, i.e. albedo, normal and lighting. To solve this ill-posed problem from single image, state-of-the-art methods in shape from shading mostly resort to supervised training on all the components on either synthetic or real datasets. Here, we propose a new self-supervised training paradigm that 1) reduces the need for full supervision on the decomposition task and 2) takes into account the relighting task. We introduce new self-supervised loss terms that leverage the consistencies between multi-lit images (image…
Understanding shading effects in images is critical for a variety of vision and graphics problems, including intrinsic image decomposition, shadow removal, image relighting, and inverse rendering. As is the case with other vision tasks, machine learning is a promising approach to understanding shading - but there is little ground truth shading data available for real-world images. We introduce Shading Annotations in the Wild (SAW), a new large-scale, public dataset of shading annotations in indoor scenes, comprised of multiple forms of shading judgments obtained via crowdsourcing, along with s…
3DPR: Single Image 3D Portrait Relight using Generative Priors
Max Planck Institute for Informatics, SIC \& VIA Research Center Saarbrücken Germany；Google Inc. San Fransisco USA；Max Planck Institute for Informatics \& SIC Saarbrücken Germany；ETH Zürich Zürich Switzerland；Harvard University Cambridge USA；Massachusetts Institute of Technology Cambridge USA；Google Inc. Zürich Switzerland；Max Planck Institute for Informatics, SIC \& VIA Research Center；Google Inc.；Max Planck Institute for Informatics \& SIC；ETH Zürich；Harvard University；Massachusetts Institute of Technology
University of Science and Technology of China（中国科学技术大学）
Most modern face completion approaches adopt an autoencoder or its variants to restore missing regions in face images. Encoders are often utilized to learn powerful representations that play an important role in meeting the challenges of sophisticated learning tasks. Specifically, various kinds of masks are often presented in face images in the wild, forming complex patterns, especially in this hard period of COVID-19. It's difficult for encoders to capture such powerful representations under this complex situation. To address this challenge, we propose a self-supervised Siamese inference netw…
Faculty of Information Technology, University of Science, VNU-HCM（越南国家大学胡志明市分校科学大学信息技术学院）
Kinship verification and kinship retrieval are emerging tasks in computer vision. Kinship verification aims at determining whether two facial images are from related people or not, while kinship retrieval is the task of retrieving possible related facial images to a person from a gallery of images. They introduce unique challenges because of the hidden relations and features that carry inherent characteristics between the facial images. We employ 3 methods, FaceNet, Siamese VGG-Face, and a combination of FaceNet and VGG-Face models as feature extractors, to achieve the 9th standing for kinship…
MPI for Informatics, SIC（马克斯-普朗克信息学研究所（MPI for Informatics, SIC））；InterDigital R&I（InterDigital研发（InterDigital R&I））；University College London（伦敦大学学院（University College London））；IST Austria（奥地利科学技术研究所（IST Austria））；Harvard University（哈佛大学（Harvard University））；MIT CSAIL（麻省理工学院计算机科学与人工智能实验室（MIT CSAIL））
Photorealistic editing of portraits is a challenging task as humans are very sensitive to inconsistencies in faces. We present an approach for high-quality intuitive editing of the camera viewpoint and scene illumination in a portrait image. This requires our method to capture and control the full reflectance field of the person in the image. Most editing approaches rely on supervised learning using training data captured with setups such as light and camera stages. Such datasets are expensive to acquire, not readily available and do not capture all the rich variations of in-the-wild portrait …
