Halloween 3d face changing photo5/20/2023 ![]() Reconstructions can be done on both Windows and Linux. ![]() ResNet-50 is used as backbone network to achieve over 50 fps (on GTX 1080) for reconstructions. We conduct an experiment on AFLW_2000 dataset (NME) to evaluate the performance, as shown in the table below:įaces are represented with Basel Face Model 2009, which is easy for further manipulations (e.g expression transfer). It provides face pose estimation and 68 facial landmarks which are useful for other tasks. Our method aligns reconstruction faces with input images. The method can provide reasonable results under extreme conditions such as large pose and occlusions. Scene illumination is also disentangled to generate a pure albedo. The method produces high fidelity face textures meanwhile preserves identity information of input images. (Please refer to our paper for more details about these results) Quantitative evaluations (shape errors in mm) on several benchmarks show its state-of-the-art performance: Method The method reconstructs faces with high accuracy. It achieves state-of-the-art performance on multiple datasets such as FaceWarehouse, MICC Florence and BU-3DFE. It is fast, accurate, and robust to pose and occlussions. The method enforces a hybrid-level weakly-supervised training for CNN-based 3D face reconstruction. Tong, Accurate 3D Face Reconstruction with Weakly-Supervised Learning: From Single Image to Image Set, IEEE Computer Vision and Pattern Recognition Workshop (CVPRW) on Analysis and Modeling of Faces and Gestures (AMFG), 2019. This is a tensorflow implementation of the following paper: This repo will not be maintained in future. ***: A PyTorch implementation which has much better performance and is much easier to use is available now. Accurate 3D Face Reconstruction with Weakly-Supervised Learning: From Single Image to Image Set
0 Comments
Leave a Reply. |