Adaptive 3D Face Reconstruction from Unconstrained Photo Collections

By Joseph Roth, Y. Tong, Xiaoming Liu. 2017

176 citations recorded by the source corpus.

Subfields: Computational Mechanics, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Mechanics

In Computer Vision and Pattern Recognition

Citations in the corpus

A Hybrid Model for Identity Obfuscation by Face Replacement (2018, background): “Such parametric models can act as strong regularizers for 3D face reconstruction problems, and have been widely used in optimization based [5,14,12,15,16,17] and learningbased [18,19,20,21,22,23] settings.”
3D Face Reconstruction by Learning from Synthetic Data (2016, background): “This is sufficient for some tasks, such as in [28], while other require further refinement as applied in [22] for facial reconstruction from multiple images.”
Learning Detailed Face Reconstruction from a Single Image (2016, methodology): “(e) is a reconstruction of the same subject, based on 100 different images recovered with the method proposed in [36].”
Text-based editing of talking-head video (2019, methodology): “There is a large body of work on reconstructing facial geometry and appearance from a single image using optimization methods[Fyffe et al. 2014; Garrido et al. 2016; Ichim et al. 2015; Kemelmacher-Shlizerman 2013; Roth et al. 2017; Shi et al. 2014; Suwajanakorn et al. 2017; Thies et al. 2016].”
Large-Pose Face Alignment via CNN-Based Dense 3D Model Fitting (2016, background): “…such as eye corners, nose tip, etc. Improving the face alignment accuracy is beneficial for many computer vision tasks related to facial analysis, because it is used as a prerequisite for these tasks, e.g., face recognition [28], 3D face reconstruction [20, 21] and face de-identification [10].”
SfSNet: Learning Shape, Reflectance and Illuminance of Faces 'in the Wild' (2017, background): “Recent works used facial priors to reconstruct shape from a single image [14, 13, 6, 26] or from multiple images [25].”
Accurate 3D Face Reconstruction With Weakly-Supervised Learning: From Single Image to Image Set (2019, methodology): “Our method is also related to traditional methods working on unconstrained photo collections [27, 47, 40, 41].”
Extreme 3D Face Reconstruction: Seeing Through Occlusions (2017, methodology): “Before reviewing single view 3D reconstruction methods, we note that some methods were developed for multi view reconstruction [18, 27, 41, 42, 46].”
Face Alignment in Full Pose Range: A 3D Total Solution (2015, background): “tinct task is 3D face reconstruction [15], [50], [51], [52], which recovers a 3D face from given 2D landmarks.”
Towards Large-Pose Face Frontalization in the Wild (2017, methodology): “rance to match an input image as closely as possible. Subsequently, the recovered shape and appearance can be used to generate a face image under novel viewpoints. Many 3D face reconstruction methods [22, 43, 23] build upon this direction by improving speed or accuracy. Deep learning has made inroad”
MoFA: Model-Based Deep Convolutional Face Autoencoder for Unsupervised Monocular Reconstruction (2017, methodology): “Previous approaches have tackled this challenging problem using calibrated multi-view data or uncalibrated photo collections [26, 45].”
Self-Supervised Multi-level Face Model Learning for Monocular Reconstruction at Over 250 Hz (2017, background): “Optimization-based Approaches: Many approaches for monocular face reconstruction [50], reconstruction based on image collections [51], and the estimation of highquality 3D face rigs [26] are based on energy optimization.”
Extreme 3D Face Reconstruction: Looking Past Occlusions (2017, methodology): “h resolution images are available on our project page. 1 2. Related work Before reviewing single view 3D reconstruction methods, we note that some methods were developed for multi view reconstruction [19,28,42,43,47]. With multiple views, better guarantees can be made on reconstruction accuracy and ”
Face anti-spoofing using patch and depth-based CNNs (2017, background): “Specifically, for face images, face reconstruction from one image [20, 21] or multiple images [37, 38] can also be viewed as one way for depth estimation.”
Large Pose 3D Face Reconstruction from a Single Image via Direct Volumetric CNN Regression (2017, methodology): “More recently, the state-of-the-art method of [21] that produces the average (neutral) 3D face from a collection of personal photos, firstly performs landmark detection, then fits a 3DMM using a sparse set of points, then solves an optimization problem similar to the one in [11], then performs surfa”
Regressing Robust and Discriminative 3D Morphable Models with a Very Deep Neural Network (2016, background): “Before listing them, it is important to mention recent multi image methods which use image sets for reconstruction (e.g., [23, 28, 32, 33, 36]).”
Learning an animatable detailed 3D face model from in-the-wild images (2020, methodology): “…al. 2015; Jeni et al. 2015; Shi et al. 2014; Suwajanakorn et al. 2014], RGB-D data [Li et al. 2013; Thies et al. 2015; Weise et al. 2011] or subject-specific image collections [Kemelmacher-Shlizerman and Seitz 2011; Roth et al. 2016], our main focus is on methods that use only a single RGB image.”
X2Face: A network for controlling face generation by using images, audio, and pose codes (2018, background): “fitting of 3DMMs by including high level details [34, 41], taking into account additional images [33] or 3D scans [4], or by learning 3DMM parameters directly from RGB data without ground truth labels [2, 39].”
LRM: Large Reconstruction Model for Single Image to 3D (2023, unknown): “Leveraging 3D templates (Roth et al., 2016; Goel et al., 2020; Kanazawa et al., 2018; Kulkarni et al., 2020), semantics (Li et al., 2020), and poses (Bogo et al., 2016; Novotny et al., 2019) as shape priors have also been widely studied in category-specific reconstruction.”