In 2000, the Cohn-Kanade (CK) database was released for the purpose of promoting research into automatically detecting individual facial expressions. Since then, the CK database has become one of the most widely used test-beds for algorithm development and evaluation. During this period, three limitations have become apparent: 1) While AU codes are well validated, emotion labels are not, as they refer to what was requested rather than what was actually performed, 2) The lack of a common performance metric against which to evaluate new algorithms, and 3) Standard protocols for common databases have not emerged. As a consequence, the CK database has been used for both AU and emotion detection (even though labels for the latter have not been validated), comparison with benchmark algorithms is missing, and use of random subsets of the original database makes meta-analyses difficult. To address these and other concerns, we present the Extended Cohn-Kanade (CK+) database. The number of sequences is increased by 22% and the number of subjects by 27%. The target expression for each sequence is fully FACS coded and emotion labels have been revised and validated. In addition to this, non-posed sequences for several types of smiles and their associated metadata have been added. We present baseline results using Active Appearance Models (AAMs) and a linear support vector machine (SVM) classifier using a leave-one-out subject cross-validation for both AU and emotion detection for the posed data. The emotion and AU labels, along with the extended image data and tracked landmarks will be made available July 2010.

/pdf/the-extended-cohn-kanade-dataset-ck-a-complete-dataset-for-3ui317eamh.pdf

The Extended Cohn-Kanade Dataset (CK+): A complete dataset for action unit and emotion-specified expression

This paper addresses the problem of Face Alignment for a single image. We show how an ensemble of regression trees can be used to estimate the face's landmark positions directly from a sparse subset of pixel intensities, achieving super-realtime performance with high quality predictions. We present a general framework based on gradient boosting for learning an ensemble of regression trees that optimizes the sum of square error loss and naturally handles missing or partially labelled data. We show how using appropriate priors exploiting the structure of image data helps with efficient feature selection. Different regularization strategies and its importance to combat overfitting are also investigated. In addition, we analyse the effect of the quantity of training data on the accuracy of the predictions and explore the effect of data augmentation using synthesized data.

/pdf/one-millisecond-face-alignment-with-an-ensemble-of-3lkmibejq7.pdf

One Millisecond Face Alignment with an Ensemble of Regression Trees

We present a method that achieves state-of-the-art results for synthesizing novel views of complex scenes by optimizing an underlying continuous volumetric scene function using a sparse set of input views. Our algorithm represents a scene using a fully-connected (non-convolutional) deep network, whose input is a single continuous 5D coordinate (spatial location $(x,y,z)$ and viewing direction $(\theta, \phi)$) and whose output is the volume density and view-dependent emitted radiance at that spatial location. We synthesize views by querying 5D coordinates along camera rays and use classic volume rendering techniques to project the output colors and densities into an image. Because volume rendering is naturally differentiable, the only input required to optimize our representation is a set of images with known camera poses. We describe how to effectively optimize neural radiance fields to render photorealistic novel views of scenes with complicated geometry and appearance, and demonstrate results that outperform prior work on neural rendering and view synthesis. View synthesis results are best viewed as videos, so we urge readers to view our supplementary video for convincing comparisons.

NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis

We present a unified model for face detection, pose estimation, and landmark estimation in real-world, cluttered images. Our model is based on a mixtures of trees with a shared pool of parts; we model every facial landmark as a part and use global mixtures to capture topological changes due to viewpoint. We show that tree-structured models are surprisingly effective at capturing global elastic deformation, while being easy to optimize unlike dense graph structures. We present extensive results on standard face benchmarks, as well as a new “in the wild” annotated dataset, that suggests our system advances the state-of-the-art, sometimes considerably, for all three tasks. Though our model is modestly trained with hundreds of faces, it compares favorably to commercial systems trained with billions of examples (such as Google Picasa and face.com).

Face detection, pose estimation, and landmark localization in the wild

Many computer vision problems (e.g., camera calibration, image alignment, structure from motion) are solved through a nonlinear optimization method. It is generally accepted that 2nd order descent methods are the most robust, fast and reliable approaches for nonlinear optimization of a general smooth function. However, in the context of computer vision, 2nd order descent methods have two main drawbacks: (1) The function might not be analytically differentiable and numerical approximations are impractical. (2) The Hessian might be large and not positive definite. To address these issues, this paper proposes a Supervised Descent Method (SDM) for minimizing a Non-linear Least Squares (NLS) function. During training, the SDM learns a sequence of descent directions that minimizes the mean of NLS functions sampled at different points. In testing, SDM minimizes the NLS objective using the learned descent directions without computing the Jacobian nor the Hessian. We illustrate the benefits of our approach in synthetic and real examples, and show how SDM achieves state-of-the-art performance in the problem of facial feature detection. The code is available at www.humansensing.cs. cmu.edu/intraface.

/pdf/supervised-descent-method-and-its-applications-to-face-assiism8su.pdf

Supervised Descent Method and Its Applications to Face Alignment

Deformable model fitting has been actively pursued in the computer vision community for over a decade. As a result, numerous approaches have been proposed with varying degrees of success. A class of approaches that has shown substantial promise is one that makes independent predictions regarding locations of the model's landmarks, which are combined by enforcing a prior over their joint motion. A common theme in innovations to this approach is the replacement of the distribution of probable landmark locations, obtained from each local detector, with simpler parametric forms. In this work, a principled optimization strategy is proposed where nonparametric representations of these likelihoods are maximized within a hierarchy of smoothed estimates. The resulting update equations are reminiscent of mean-shift over the landmarks but with regularization imposed through a global prior over their joint motion. Extensions to handle partial occlusions and reduce computational complexity are also presented. Through numerical experiments, this approach is shown to outperform some common existing methods on the task of generic face fitting.

/pdf/deformable-model-fitting-by-regularized-landmark-mean-shift-2bk7fb79o2.pdf

Deformable Model Fitting by Regularized Landmark Mean-Shift

Recent advances in image-based 3D human shape estimation have been driven by the significant improvement in representation power afforded by deep neural networks. Although current approaches have demonstrated the potential in real world settings, they still fail to produce reconstructions with the level of detail often present in the input images. We argue that this limitation stems primarily form two conflicting requirements; accurate predictions require large context, but precise predictions require high resolution. Due to memory limitations in current hardware, previous approaches tend to take low resolution images as input to cover large spatial context, and produce less precise (or low resolution) 3D estimates as a result. We address this limitation by formulating a multi-level architecture that is end-to-end trainable. A coarse level observes the whole image at lower resolution and focuses on holistic reasoning. This provides context to an fine level which estimates highly detailed geometry by observing higher-resolution images. We demonstrate that our approach significantly outperforms existing state-of-the-art techniques on single image human shape reconstruction by fully leveraging 1k-resolution input images.

/pdf/pifuhd-multi-level-pixel-aligned-implicit-function-for-high-2sf1hg8i1x.pdf

PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution 3D Human Digitization

Deformable model fitting has been actively pursued in the computer vision community for over a decade As a result, numerous approaches have been proposed with varying degrees of success A class of approaches that has shown substantial promise is one that makes independent predictions regarding locations of the model's landmarks, which are combined by enforcing a prior over their joint motion A common theme in innovations to this approach is the replacement of the distribution of probable landmark locations, obtained from each local detector, with simpler parametric forms This simplification substitutes the true objective with a smoothed version of itself, reducing sensitivity to local minima and outlying detections In this work, a principled optimization strategy is proposed where a nonparametric representation of the landmark distributions is maximized within a hierarchy of smoothed estimates The resulting update equations are reminiscent of mean-shift but with a subspace constraint placed on the shape's variability This approach is shown to outperform other existing methods on the task of generic face fitting

/pdf/face-alignment-through-subspace-constrained-mean-shifts-127lcxsyts.pdf

Face alignment through subspace constrained mean-shifts

Modeling and rendering of dynamic scenes is challenging, as natural scenes often contain complex phenomena such as thin structures, evolving topology, translucency, scattering, occlusion, and biological motion. Mesh-based reconstruction and tracking often fail in these cases, and other approaches (e.g., light field video) typically rely on constrained viewing conditions, which limit interactivity. We circumvent these difficulties by presenting a learning-based approach to representing dynamic objects inspired by the integral projection model used in tomographic imaging. The approach is supervised directly from 2D images in a multi-view capture setting and does not require explicit reconstruction or tracking of the object. Our method has two primary components: an encoder-decoder network that transforms input images into a 3D volume representation, and a differentiable ray-marching operation that enables end-to-end training. By virtue of its 3D representation, our construction extrapolates better to novel viewpoints compared to screen-space rendering techniques. The encoder-decoder architecture learns a latent representation of a dynamic scene that enables us to produce novel content sequences not seen during training. To overcome memory limitations of voxel-based representations, we learn a dynamic irregular grid structure implemented with a warp field during ray-marching. This structure greatly improves the apparent resolution and reduces grid-like artifacts and jagged motion. Finally, we demonstrate how to incorporate surface-based representations into our volumetric-learning framework for applications where the highest resolution is required, using facial performance capture as a case in point.

Jason Saragih

Papers

The Extended Cohn-Kanade Dataset (CK+): A complete dataset for action unit and emotion-specified expression

Deformable Model Fitting by Regularized Landmark Mean-Shift

PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution 3D Human Digitization

Face alignment through subspace constrained mean-shifts

Neural volumes: learning dynamic renderable volumes from images