Recognizing facial action units (AUs) is important for situation analysis and automated video annotation. Previous work has emphasized face tracking and registration and the choice of features classifiers. Relatively neglected is the effect of imbalanced data for action unit detection. While the machine learning community has become aware of the problem of skewed data for training classifiers, little attention has been paid to how skew may bias performance metrics. To address this question, we conducted experiments using both simulated classifiers and three major databases that differ in size, type of FACS coding, and degree of skew. We evaluated influence of skew on both threshold metrics (Accuracy, F-score, Cohen's kappa, and Krippendorf's alpha) and rank metrics (area under the receiver operating characteristic (ROC) curve and precision-recall curve). With exception of area under the ROC curve, all were attenuated by skewed distributions, in many cases, dramatically so. While ROC was unaffected by skew, precision-recall curves suggest that ROC may mask poor performance. Our findings suggest that skew is a critical factor in evaluating performance metrics. To avoid or minimize skew-biased estimates of performance, we recommend reporting skew-normalized scores along with the obtained ones.

To enable real-time, person-independent 3D registration from 2D video, we developed a 3D cascade regression approach in which facial landmarks remain invariant across pose over a range of approximately 60 degrees. From a single 2D image of a person's face, a dense 3D shape is registered in real time for each frame. The algorithm utilizes a fast cascade regression framework trained on high-resolution 3D face-scans of posed and spontaneous emotion expression. The algorithm first estimates the location of a dense set of markers and their visibility, then reconstructs face shapes by fitting a part-based 3D model. Because no assumptions are required about illumination or surface properties, the method can be applied to a wide range of imaging conditions that include 2D video and uncalibrated multi-view video. The method has been validated in a battery of experiments that evaluate its precision of 3D reconstruction and extension to multi-view reconstruction. Experimental findings strongly support the validity of real-time, 3D registration and reconstruction from 2D video. The software is available online at http://zface.org.

The field of Automatic Facial Expression Analysis has grown rapidly in recent years. However, despite progress in new approaches as well as benchmarking efforts, most evaluations still focus on either posed expressions, near-frontal recordings, or both. This makes it hard to tell how existing expression recognition approaches perform under conditions where faces appear in a wide range of poses (or camera views), displaying ecologically valid expressions. The main obstacle for assessing this is the availability of suitable data, and the challenge proposed here addresses this limitation. The FG 2017 Facial Expression Recognition and Analysis challenge (FERA 2017) extends FERA 2015 to the estimation of Action Units occurrence and intensity under different camera views. In this paper we present the third challenge in automatic recognition of facial expressions, to be held in conjunction with the 12th IEEE conference on Face and Gesture Recognition, May 2017, in Washington, United States. Two sub-challenges are defined: the detection of AU occurrence, and the estimation of AU intensity. In this work we outline the evaluation protocol, the data used, and the results of a baseline method for both sub-challenges.

Abstract-Most work in automatic facial expression analysis seeks to detect discrete facial actions. Yet, the meaning and function of facial actions often depends in part on their intensity. We propose a part-based, sparse representation for automated measurement of continuous variation in AU intensity. We evaluated its effectiveness in two publically available databases, CK+ and the soon to be released Binghamton high-resolution spontaneous 3D dyadic facial expression database. The former consists of posed facial expressions and ordinal level intensity (absent, low, and high). The latter consists of spontaneous facial expression in response to diverse, well-validated emotion inductions, and 6 ordinal levels of AU intensity.In a preliminary test, we started from discrete emotion labels and ordinal-scale intensity annotation in the CK+ dataset. The algorithm achieved state-of-the-art performance. These preliminary results supported the utility of the part-based, sparse representation. Second, we applied the algorithm to the more demanding task of continuous AU intensity estimation in spontaneous facial behavior in the Binghamton database. Manual 6-point ordinal coding and continuous measurement were highly consistent. Visual analysis of the overlay of continuous measurement by the algorithm and manual ordinal coding strongly supported the representational power of the proposed method to smoothly interpolate across the full range of AU intensity.

No abstract