Online modeling for realtime facial animation

By Sofien Bouaziz, Yangang Wang, Mark Pauly. 2013

295 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Control and Systems Engineering

How researchers see this paper

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

In Computer Vision and Pattern Recognition

In Control and Systems Engineering

Citations in the corpus

Generating 3D faces using Convolutional Mesh Autoencoders (2018, methodology): “The linear expression basis vectors are either computed using PCA [1,7,28,41,47], or are manually deﬁned using linear blendshapes (e.g. [42,27,6]).”
Reconstruction of Personalized 3D Face Rigs from Monocular Video (2016, methodology): “Some methods capture facial performances [Weise et al. 2011] and person-specific corrective shapes from RGB-D data [Bouaziz et al. 2013; Li et al. 2013], whereas our approach only requires monocular RGB video.”
Nerfies: Deformable Neural Radiance Fields (2020, background): “Domain-Specific Modeling: Many reconstruction methods use domain-specific knowledge to model the shape and appearance of categories with limited topological variation, such as faces [5, 7, 9], human bodies [30, 58], and animals [12, 64].”
Total Moving Face Reconstruction (2014, background): “Instead of requiring the person to be scanned in the lab or participate in the reconstruction process (as many other methods require [24,18,10,41,33]), we leverage whatever existing imagery is available online or in personal photo collections.”
Practice and Theory of Blendshape Facial Models (2014, background): “[BWP13] combine a PCA model of identity with a blend-shape model of expressions obtained through deformation transfer from a generic template model.”
Dynamic Neural Radiance Fields for Monocular 4D Facial Avatar Reconstruction (2020, methodology): “Some methods [9, 23, 8, 17] also focused on corrective shapes [8], dynamically adapting the blendshape basis [23] or applied non-rigid mesh deformation [9] to compensate for the coarse geometry of the morphable model.”
Automatic acquisition of high-fidelity facial performances using monocular videos (2014, methodology): “Our system is also relevant to recent successes in tracking 3D facial expression using a single RGBD camera such as Microsoft Kinect or time-of-flight (TOF) cameras [Weise et al. 2011; Bouaziz et al. 2013; Li et al. 2013].”
Displaced dynamic expression regression for real-time facial tracking and animation (2014, methodology): “The recent advent of RGBD cameras has enabled a number of depth-based facial tracking methods [Weise et al. 2011; Baltrušaitis et al. 2012; Bouaziz et al. 2013; Li et al. 2013].”
Real-time expression transfer for facial reenactment (2015, methodology): “Real-time performance on dense RGB-D input has recently been achieved by tracking a personalized blend shape model [Weise et al. 2011; Li et al. 2013; Bouaziz et al. 2013; Hsieh et al. 2015], or by the deformation of a face template mesh in an as-rigid-as-possible framework [Zollhöfer et al. 2014].”
paGAN (2018, unknown): “Nevertheless, these dynamic face models are suitable for single-view face modeling and tracking applications [Blanz and Vetter 1999; Bouaziz et al. 2013; Cao et al. 2014, 2016; Garrido et al. 2016; Hsieh et al. 2015; Hu et al. 2017; Li et al. 2013; Saito et al. 2016, 2017; Thies et al. 2015, 2016a; ”
Face2Face: Real-Time Face Capture and Reenactment of RGB Videos (2016, background): “No additional smoothness term as in [27, 4] is needed, since the blendshape model implicitly restricts the result to plausible shapes and guarantees smoothness.”
3D Morphable Face Models—Past, Present, and Future (2019, unknown): “During tracking, Li et al. [2013] use an adaptive PCA model that combines the person-specific blendshapes with additional corrective basis vectors that are successively updated, and Bouaziz et al. [2013] optimizes for corrective deformation fields (Equation 4).”
Self-Supervised Multi-level Face Model Learning for Monocular Reconstruction at Over 250 Hz (2017, background): “[12] introduced medium-scale shape correctives based on manifold harmonics [65].”