人脸序列图象的头部运动参数估计

在人脸序列的图象编码中 ,由于模型基编码方法可以获得高的主观图象质量和低的码率 ,因而受到广泛重视 .但是 ,其运动参数的可靠估计还是一个难点 .为此 ,该文分析了头部运动的特点 ,并把它分为头部刚体运动、脸部表情的简单运动和脸部表情复杂运动 3种形式 .其中 ,提取头部刚体运动参数利用了基于特征点对的运动参数估计算法 ,并提出了一个线性的实现方法 ;文中还指出提高运动参数估计的精度在于选择合适的特征点和建立一个和特定人脸相一致的三维线框模型 ;另外 ,还为脸部表情的简单运动建立了形变矩阵 ;最后给出了用面积误差函数评价的运动参数估计误差 .     

一种基于区域分割的表情鲁棒三维人脸识别方法

为了克服表情变化致使三维人脸识别性能不佳的问题,提出基于鼻尖点区域分割的表情鲁棒三维人脸识别方法。首先,根据表情对人脸影响具有区域性的特点,提出仅依赖鼻尖点的表情不变区域（刚性区域）和表情易变（非刚性区域）划分方法;然后针对表情不变区域和表情易变区域使用不同的特征描述方式并计算匹配相似度;最后将表情不变区域和表情易变的相似度进行加权融合实现最终身份识别。提出的方法分别在FRGC v2.0和自建WiseFace表情人脸数据库上达到98.52%和99.01%的rank 1识别率,证明该方法对表情变化具有较强的鲁棒性。     

A Unified Probabilistic Framework for Spontaneous Facial Action Modeling and Understanding

Facial expression is a natural and powerful means of human communication. Recognizing spontaneous facial actions, however, is very challenging due to subtle facial deformation, frequent head movements, and ambiguous and uncertain facial motion measurements. Because of these challenges, current research in facial expression recognition is limited to posed expressions and often in frontal view. A spontaneous facial expression is characterized by rigid head movements and nonrigid facial muscular movements. More importantly, it is the coherent and consistent spatiotemporal interactions among rigid and nonrigid facial motions that produce a meaningful facial expression. Recognizing this fact, we introduce a unified probabilistic facial action model based on the Dynamic Bayesian network (DBN) to simultaneously and coherently represent rigid and nonrigid facial motions, their spatiotemporal dependencies, and their image measurements. Advanced machine learning methods are introduced to learn the model based on both training data and subjective prior knowledge. Given the model and the measurements of facial motions, facial action recognition is accomplished through probabilistic inference by systematically integrating visual measurements with the facial action model. Experiments show that compared to the state-of-the-art techniques, the proposed system yields significant improvements in recognizing both rigid and nonrigid facial motions, especially for spontaneous facial expressions.     

Non-Frontal Facial Expression Recognition Using a Depth-Patch Based Deep Neural Network

The challenge of coping with non-frontal head poses during facial expression recognition results in considerable reduction of accuracy and robustness when capturing expressions that occur during natural communications. In this paper, we attempt to recognize facial expressions under poses with large rotation angles from 2D videos. A depth-patch based 4D expression representation model is proposed. It was reconstructed from 2D dynamic images for delineating continuous spatial changes and temporal context under non-frontal cases. Furthermore, we present an effective deep neural network classifier, which can accurately capture pose-variant expression features from the depth patches and recognize non-frontal expressions. Experimental results on the BU-4DFE database show that the proposed method achieves a high recognition accuracy of 86.87% for non-frontal facial expressions within a range of head rotation angle of up to 52°, outperforming existing methods. We also present a quantitative analysis of the components contributing to the performance gain through tests on the BU-4DFE and Multi-PIE datasets.     

Statistical learning based facial animation

To synthesize real-time and realistic facial animation, we present an effective algorithm which combines image- and geometry-based methods for facial animation simulation. Considering the numerous motion units in the expression coding system, we present a novel simplified motion unit based on the basic facial expression, and construct the corresponding basic action for a head model. As image features are difficult to obtain using the performance driven method, we develop an automatic image feature recognition method based on statistical learning, and an expression image semi-automatic labeling method with rotation invariant face detection, which can improve the accuracy and efficiency of expression feature identification and training. After facial animation redirection, each basic action weight needs to be computed and mapped automatically. We apply the blend shape method to construct and train the corresponding expression database according to each basic action, and adopt the least squares method to compute the corresponding control parameters for facial animation. Moreover, there is a pre-integration of diffuse light distribution and specular light distribution based on the physical method, to improve the plausibility and efficiency of facial rendering. Our work provides a simplification of the facial motion unit, an optimization of the statistical training process and recognition process for facial animation, solves the expression parameters, and simulates the subsurface scattering effect in real time. Experimental results indicate that our method is effective and efficient, and suitable for computer animation and interactive applications.     

Reconstruction of non-rigid 3D shapes from stereo-motion

Several non-rigid structure from motion methods have been proposed so far in order to recover both the motion and the non-rigid structure of an object. However, these monocular algorithms fail to give reliable 3D shape estimates when the overall rigid motion of the sequence is small. Aiming to overcome this limitation, in this paper we propose a novel approach for the 3D Euclidean reconstruction of deformable objects observed by an uncalibrated stereo rig. Using a stereo setup drastically improves the 3D model estimation when the observed 3D shape is mostly deforming without undergoing strong rigid motion. Our approach is based on the following steps. Firstly, the stereo system is automatically calibrated and used to compute metric rigid structures from pairs of views. Afterwards, these 3D shapes are aligned to a reference view using a RANSAC method in order to compute the mean shape of the object and to select the subset of points which have remained rigid throughout the sequence. The selected rigid points are then used to compute frame-wise shape registration and to robustly extract the motion parameters from frame to frame. Finally, all this information is used as initial estimates of a non-linear optimization which allows us to refine the initial solution and also to recover the non-rigid 3D model. Exhaustive results on synthetic and real data prove the performance of our proposal estimating motion, non-rigid models and stereo camera parameters even when there is no rigid motion in the original sequence.     

刚性区域特征点的3维人脸识别

目的 针对3维人脸识别中存在表情变化的问题,提出了一种基于刚性区域特征点的3维人脸识别方法。方法 该方法首先在人脸纹理图像上提取人脸图像的特征点,并删除非刚性区域内的特征点,然后根据采样点的序号,在人脸空间几何信息上得到人脸图像特征点的3维几何信息,并建立以特征点为中心的刚性区域内的子区域,最后以子区域为局部特征进行人脸识别测试,得到不同子区域对人脸识别的贡献,并以此作为依据对人脸识别的结果进行加权统计。结果 在FRGC v2.0的3维人脸数据库上进行实验测试,该方法的识别准确率为98.5%,当错误接受率（FAR）为0.001时的验证率为99.2%,结果表明,该方法对非中性表情下的3维人脸识别具有很好的准确性。结论 该方法可以有效克服表情变化对3维人脸识别的影响,同时对3维数据中存在的空洞和尖锐噪声等因素具有较好的鲁棒性,对提高3维人脸识别性能具有重要意义。     

基于双相机捕获面部表情及人体姿态生成三维虚拟人动画

为了生成表情丰富、动作流畅的三维虚拟人动画,提出了一种基于双相机同步捕获面部表情及人体姿态生成三维虚拟人动画的方法。首先,采用传输控制协议（TCP）网络时间戳方法实现双相机时间同步,采用张正友标定法实现双相机空间同步。然后,利用双相机分别采集面部表情和人体姿态。采集面部表情时,提取图像的2D特征点,利用这些2D特征点回归计算得到面部行为编码系统（FACS）面部行为单元,为实现表情动画做准备;以标准头部3D坐标值为基准,根据相机内参,采用高效n点投影（EPnP）算法实现头部姿态估计;之后将面部表情信息和头部姿态估计信息进行匹配。采集人体姿态时,利用遮挡鲁棒姿势图（ORPM）方法计算人体姿态,输出每个骨骼点位置、旋转角度等数据。最后,在虚幻引擎4（UE4）中使用建立的虚拟人体三维模型来展示数据驱动动画的效果。实验结果表明,该方法能够同步捕获面部表情及人体姿态,而且在实验测试中的帧率达到20 fps,能实时生成自然真实的三维动画。     

Expression-robust 3D face recognition based on feature-level fusion and feature-region fusion

3D face shape is essentially a non-rigid free-form surface, which will produce non-rigid deformation under expression variations. In terms of that problem, a promising solution named Coherent Point Drift (CPD) non-rigid registration for the non-rigid region is applied to eliminate the influence from the facial expression while guarantees 3D surface topology. In order to take full advantage of the extracted discriminative feature of the whole face under facial expression variations, the novel expression-robust 3D face recognition method using feature-level fusion and feature-region fusion is proposed. Furthermore, the Principal Component Analysis and Linear Discriminant Analysis in combination with Rotated Sparse Regression (PL-RSR) dimensionality reduction method is presented to promote the computational efficiency and provide a solution to the curse of dimensionality problem, which benefit the performance optimization. The experimental evaluation indicates that the proposed strategy has achieved the rank-1 recognition rate of 97.91 % and 96.71 % based on Face Recognition Grand Challenge (FRGC) v2.0 and Bosphorus respectively, which means the proposed approach outperforms state-of-the-art approach.     

结合像素模式和特征点模式的实时表情识别

目的 目前2D表情识别方法对于一些混淆性较高的表情识别率不高并且容易受到人脸姿态、光照变化的影响,利用RGBD摄像头Kinect获取人脸3D特征点数据,提出了一种结合像素2D特征和特征点3D特征的实时表情识别方法。方法 首先,利用3种经典的LBP（局部二值模式）、Gabor滤波器、HOG（方向梯度直方图）提取了人脸表情2D像素特征,由于2D像素特征对于人脸表情描述能力的局限性,进一步提取了人脸特征点之间的角度、距离、法向量3种3D表情特征,以对不同表情的变化情况进行更加细致地描述。为了提高算法对混淆性高的表情识别能力并增加鲁棒性,将2D像素特征和3D特征点特征分别训练了3组随机森林模型,通过对6组随机森林分类器的分类结果加权组合,得到最终的表情类别。结果 在3D表情数据集Face3D上验证算法对9种不同表情的识别效果,结果表明结合2D像素特征和3D特征点特征的方法有利于表情的识别,平均识别率达到了84.7%,高出近几年提出的最优方法4.5%,而且相比单独地2D、3D融合特征,平均识别率分别提高了3.0%和5.8%,同时对于混淆性较强的愤怒、悲伤、害怕等表情识别率均高于80%,实时性也达到了10~15帧/s。结论 该方法结合表情图像的2D像素特征和3D特征点特征,提高了算法对于人脸表情变化的描述能力,而且针对混淆性较强的表情分类,对多组随机森林分类器的分类结果加权平均,有效地降低了混淆性表情之间的干扰,提高了算法的鲁棒性。实验结果表明了该方法相比普通的2D特征、3D特征等对于表情的识别不仅具有一定的优越性,同时还能保证算法的实时性。     

Model-free non-rigid head pose tracking by joint shape and pose estimation

Head pose estimation under non-rigid face movement is particularly useful in applications relating to eye-gaze tracking in less constrained scenarios, where the user is allowed to move naturally during tracking. Existing vision-based head pose estimation methods often require accurate initialisation and tracking of specific facial landmarks, while methods that handle non-rigid face deformations typically necessitate a preliminary training phase prior to head pose estimation. In this paper, we propose a method to estimate the head pose in real-time from the trajectories of a set of feature points spread randomly over the face region, without requiring a training phase or model-fitting of specific facial features. Conversely, our method exploits the 3-dimensional shape of the surface of interest, recovered via shape and motion factorisation, in combination with Kalman and particle filtering to determine the contribution of each feature point to the estimation of head pose based on a variance measure. Quantitative and qualitative results reveal the capability of our method in handling non-rigid face movement without deterioration of the head pose estimation accuracy.     

Facial expression recognition using tracked facial actions: Classifier performance analysis

In this paper, we address the analysis and recognition of facial expressions in continuous videos. More precisely, we study classifiers performance that exploit head pose independent temporal facial action parameters. These are provided by an appearance-based 3D face tracker that simultaneously provides the 3D head pose and facial actions. The use of such tracker makes the recognition pose- and texture-independent. Two different schemes are studied. The first scheme adopts a dynamic time warping technique for recognizing expressions where training data are given by temporal signatures associated with different universal facial expressions. The second scheme models temporal signatures associated with facial actions with fixed length feature vectors (observations), and uses some machine learning algorithms in order to recognize the displayed expression. Experiments quantified the performance of different schemes. These were carried out on CMU video sequences and home-made video sequences. The results show that the use of dimension reduction techniques on the extracted time series can improve the classification performance. Moreover, these experiments show that the best recognition rate can be above 90%.     

Data-driven facial expression synthesis via Laplacian deformation

Realistic talking heads have important use in interactive multimedia applications. This paper presents a novel framework to synthesize realistic facial animations driven by motion capture data using Laplacian deformation. We first capture the facial expression from a performer, then decompose the motion data into two components: the rigid movement of the head and the change of the facial expression. By making use of the local-detail preserving property of the Laplacian coordinate, we clone the captured facial expression onto a neutral 3D facial model using Laplacian deformation. We choose some expression “independent points” in the facial model as the fixed points when solving the Laplacian deformation equations. Experimental results show that our approach can synthesize realistic facial expressions in real time while preserving the facial details. We compare our method with the state-of-the-art facial expression synthesis methods to verify the advantages of our method. Our approach can be applied in real-time multimedia systems.     

Information extraction from image sequences of real-world facial expressions

Information extraction of facial expressions deals with facial-feature detection, feature tracking, and capture of the spatiotemporal relationships among features. It is a fundamental task in facial expression analysis and will ultimately determine the performance of expression recognition. For a real-world facial expression sequence, there are three challenges: (1) detection failure of some or all facial features due to changes in illumination and rapid head movement; (2) nonrigid object tracking resulting from facial expression change; and (3) feature occlusion due to out-of-plane head rotation. In this paper, a new approach is proposed to tackle these challenges. First, we use an active infrared (IR) illumination to reliably detect pupils under variable lighting conditions and head orientations. The pupil positions are then used to guide the entire information-extraction process. The simultaneous use of a global head motion constraint and Kalman filtering can robustly track individual facial features even in condition of rapid head motion and significant expression change. To handle feature occlusion, we propose a warping-based reliability propagation method. The reliable neighbor features and the spatial semantics among these features are used to detect and infer occluded features through an interframe warping transformation. Experimental results show that accurate information extraction can be achieved for video sequences with real-world facial expressions.Received: 16 August 2003, Accepted: 20 September 2004, Published online: 20 December 2004 Correspondence to: Qiang Ji     

A Video-Based Facial Motion Tracking and Expression Recognition System

We proposed a facial motion tracking and expression recognition system based on video data. By a 3D deformable facial model, the online statistical model (OSM) and cylinder head model (CHM) were combined to track 3D facial motion in the framework of particle filtering. For facial expression recognition, a fast and efficient algorithm and a robust and precise algorithm were developed. With the first, facial animation and facial expression were retrieved sequentially. After that facial animation was obtained, facial expression was recognized by static facial expression knowledge learned from anatomical analysis. With the second, facial animation and facial expression were simultaneously retrieved to increase the reliability and robustness with noisy input data. Facial expression was recognized by fusing static and dynamic facial expression knowledge, the latter of which was learned by training a multi-class expressional Markov process using a video database. The experiments showed that facial motion tracking by OSM+CHM is more pose robust than that by OSM, and the facial expression score of the robust and precise algorithm is higher than those of other state-of-the-art facial expression recognition methods.     

基于生成式对抗网络的鲁棒人脸表情识别

人们在自然情感交流中经常伴随着头部旋转和肢体动作,它们往往导致较大范围的人脸遮挡,使得人脸图像损失部分表情信息.现有的表情识别方法大多基于通用的人脸特征和识别算法,未考虑表情和身份的差异,导致对新用户的识别不够鲁棒.本文提出了一种对人脸局部遮挡图像进行用户无关表情识别的方法.该方法包括一个基于Wasserstein生成式对抗网络（Wasserstein generative adversarial net,WGAN）的人脸图像生成网络,能够为图像中的遮挡区域生成上下文一致的补全图像;以及一个表情识别网络,能够通过在表情识别任务和身份识别任务之间建立对抗关系来提取用户无关的表情特征并推断表情类别.实验结果表明,我们的方法在由CK+,Multi-PIE和JAFFE构成的混合数据集上用户无关的平均识别准确率超过了90%.在CK+上用户无关的识别准确率达到了96%,其中4.5%的性能提升得益于本文提出的对抗式表情特征提取方法.此外,在45°头部旋转范围内,本文方法还能够用于提高非正面表情的识别准确率.     

Nonrigid structure-from-motion: estimating shape and motion with hierarchical Priors

This paper describes methods for recovering time-varying shape and motion of non-rigid 3D objects from uncalibrated 2D point tracks. For example, given a video recording of a talking person, we would like to estimate the 3D shape of the face at each instant, and learn a model of facial deformation. Time-varying shape is modeled as a rigid transformation combined with a non-rigid deformation. Reconstruction is ill-posed if arbitrary deformations are allowed, and thus additional assumptions about deformations are required. We first suggest restricting shapes to lie within a low-dimensional subspace, and describe estimation algorithms. However, this restriction alone is insufficient to constrain reconstruction. To address these problems, we propose a reconstruction method using a Probabilistic Principal Components Analysis (PPCA) shape model, and an estimation algorithm that simultaneously estimates 3D shape and motion for each instant, learns the PPCA model parameters, and robustly fills-in missing data points. We then extend the model to model temporal dynamics in object shape, allowing the algorithm to robustly handle severe cases of missing data.     

Visual facial expression modeling and early predicting from 3D data via subtle feature enhancing

This work investigates a new challenging problem: how to exactly recognize facial expression captured by a high-frame rate 3D sensing as early as possible, while most works generally focus on improving the recognition rate of 2D facial expression recognition. The recognition of subtle facial expressions in their early stage is unfortunately very sensitive to noise that cannot be ignored due to their low intensity. To overcome this problem, two novel feature enhancement methods, namely, adaptive wavelet spectral subtraction method and SVM-based linear discriminant analysis, are proposed to refine subtle features of facial expressions by employing an estimated noise model or not. Experiments on a custom-made dataset built using a high-speed 3D motion capture system corroborated that the two proposed methods outperform other feature refinement methods by enhancing the discriminability of subtle facial expression features and consequently make correct recognitions earlier.     

Facial Landmark Detection: A Literature Survey

International Journal of Computer Vision - The locations of the fiducial facial landmark points around facial components and facial contour capture the rigid and non-rigid facial deformations due...