https://icon.is.tue.mpg.de/ * Home * Download * License * Login * Username + Register for other projects + Logout ICON: Implicit Clothed humans Obtained from Normals Yuliang Xiu^1, Jinlong Yang^1, Dimitrios Tzionas^1,2, Michael J. Black^1 ^1Max Planck Institute for Intelligent Systems ^2University of Amsterdam CVPR 2022 Google Colaboratory SVG HuggingFace Space Logo logo Paper Code Keynote Poster Colab HuggingFace [teaser] Introduction Video * [2022/07/30] [6874747073] Google Colab are ready to play with! * [2022/07/26] New cloth-refinement module is released, try -loop_cloth * [2022/06/13] ETH Zurich students from 3DV course create an add-on for garment-extraction. * [2022/05/16] BEV is supported as optional HPS by Yu Sun. * [2022/05/15] Training code is released, please check Training Instruction. * [2022/04/26] HybrIK (SMPL) is supported as optional HPS by Jiefeng Li. * [2022/03/05] PIXIE (SMPL-X), PARE (SMPL), PyMAF (SMPL) are all supported as optional HPS. Interactive Demo from HuggingFace Space Overview Current methods for learning realistic and animatable 3D clothed avatars need either posed 3D scans or 2D images with carefully controlled user poses. In contrast, our goal is to learn the avatar from only 2D images of people in unconstrained poses. Given a set of images, our method estimates a detailed 3D surface from each image and then combines these into an animatable avatar. Implicit functions are well suited to the first task, as they can capture details like hair or clothes. Current methods, however, are not robust to varied human poses and often produce 3D surfaces with broken or disembodied limbs, missing details, or non-human shapes. The problem is that these methods use global feature encoders that are ICON: Implicit Clothed sensitive to global pose. To address this, we humans Obtained from propose ICON ("Implicit Clothed humans Normals Obtained from Normals"), which uses local features, instead. ICON has two main modules, both of which exploit the SMPL(-X) body Google model. First, ICON infers detailed Colaboratory clothed-human normals (front/back) SVG Logo conditioned on the SMPL(-X) normals. Second, Paper Code a visibility-aware implicit surface regressor Colab produces an iso-surface of a human occupancy field. Importantly, at inference time, a feedback loop alternates between refining the SMPL(-X) mesh using the inferred clothed normals and then refining the normals. Given multiple reconstructed frames of a subject in varied poses, we use SCANimate to produce an animatable avatar from them. Evaluation on the AGORA and CAPE datasets shows that ICON outperforms the state of the art in reconstruction, even with heavily limited training data. Additionally, it is much more robust to out-of-distribution samples, e.g., in-the-wild poses/images and out-of-frame cropping. ICON takes a step towards robust 3D clothed human reconstruction from in-the-wild images. This enables creating avatars directly from video with personalized and natural pose-dependent cloth deformation. Main Contributions * Body-Guided Normal Prediction & Local-feature based implicit 3D reconstruction ICON takes, as input, an RGB image of a segmented clothed human and a SMPL body estimated from the image. The SMPL body is used to guide two of ICON's modules: one infers detailed clothed-human surface normals (front and back views), and the other infers a visibility-aware implicit surface (iso-surface of an occupancy field). [architectu] * SMPL refinement in a feedback loop Errors in the initial SMPL estimate, however, might misguide inference. Thus, at inference time, an iterative feedback loop refines SMPL (i.e. its 3D shape, pose and translation) using the inferred detailed normals, and vice versa, leading to a refined implicit shape with better 3D details. [refinement] Qualitative Results (Only Geometry) * Comparison with other state-of-the-art methods [compare] * Reconstruction on in-the-wild photos with extreme poses(GIF) [large-scale] * Reconstruction on in-the-wild photos with extreme poses (PNG) [demo] * Predicted normals on in-the-wild images with extreme poses [normal-pre] Application: create an animatable avatar from 400 images [scanimate] Human Digitization with Implicit Representation * PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human Digitization (ICCV 2019) * Shunsuke Saito, Zeng Huang, Ryota Natsume, Shigeo Morishima, Angjoo Kanazawa, Hao Li * Original PIFu. * Monocular Real-Time Volumetric Performance Capture (ECCV 2020) * Yuliang Xiu, Ruilong Li, Shunsuke Saito, Zeng Huang, Kyle Olszewski, Hao Li * Real-Time PIFu, speed++! * PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution 3D Human Digitization (CVPR 2020) * Shunsuke Saito, Tomas Simon, Jason Saragih, Hanbyul Joo * High-Resolution PIFu, details++! * PaMIR: Parametric Model-Conditioned Implicit Representation for Image-based Human Reconstruction (TPAMI2021) * Zerong Zheng, Tao Yu, Yebin Liu, Qionghai Dai * SMPL-guided PIFu, robustness++! * SCANimate: Weakly Supervised Learning of Skinned Clothed Avatar Networks (CVPR2021) * Shunsuke Saito, Jinlong Yang, Qianli Ma, and Michael J. Black * Create an avatar with pose-dependent clothing deformation from raw scans without registration. Acknowlegments We thank Yao Feng, Soubhik Sanyal, Qianli Ma, Xu Chen, Hongwei Yi, Chun-Hao Paul Huang, and Weiyang Liu for their feedback and discussions, Tsvetelina Alexiadis for her help with the AMT perceptual study, Taylor McConnell for her voice over, Benjamin Pellkofer for webpage, and Yuanlu Xu's help in comparing with ARCH and ARCH++. Special thanks to Vassilis Choutas for sharing the code of bvh-distance-queries This project has received funding from the European Union's Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No.860768 (CLIPE Project) Disclosure MJB has received research gift funds from Adobe, Intel, Nvidia, Meta/ Facebook, and Amazon. MJB has financial interests in Amazon, Datagen Technologies, and Meshcapade GmbH. While MJB was a part-time employee of Amazon during this project, his research was performed solely at, and funded solely by, the Max Planck Society. Bibtex @inproceedings{xiu2022icon, title = {{ICON}: {I}mplicit {C}lothed humans {O}btained from {N}ormals}, author = {Xiu, Yuliang and Yang, Jinlong and Tzionas, Dimitrios and Black, Michael J.}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month = {June}, year = {2022}, pages = {13296-13306} } Contact For more questions, please contact icon@tue.mpg.de For commercial licensing, please contact ps-licensing@tue.mpg.de (c) 2022 Max-Planck-Gesellschaft -Imprint-Privacy Policy-License