Eigen Light-Fields and Face Recognition Across Pose
نویسندگان
چکیده
In many face recognition tasks the pose of the probe and gallery images are different. In other cases multiple gallery or probe images may be available, each captured from a different pose. We propose a face recognition algorithm which can use any number of gallery images per subject captured at arbitrary poses, and any number of probe images, again captured at arbitrary poses. The algorithm operates by estimating the eigen light-field of the subject’s head from the input gallery or probe images. Matching between the probe and gallery is then performed using the eigen light-fields. We present results on the CMU PIE and the FERET face databases.
منابع مشابه
Face Recognition Across Pose and Illumination
The last decade has seen automatic face recognition evolve from small scale research systems to a wide range of commercial products. Driven by the FERET face database and evaluation protocol, the currently best commercial systems achieve verification accuracies comparable to those of fingerprint recognizers. In these experiments, only frontal face images taken under controlled lighting conditio...
متن کاملAn HMM-based Temporal Difference Learning with Model-Updating Capability for Visual Tracking of Human Communicational Behaviors
p. ix Conference Organizers p. x Program Committee p. xi Face Analysis Eigen Light-Fields and Face Recognition across Pose p. 3 Real-Time, Fully Automatic Upper Facial Feature Tracking p. 10 A Unified Learning Framework for Real Time Face Detection and Classification p. 16 Detection and Recognition Consistency and Coupling in Human Model Likelihoods p. 27 An Investigation into the Use of Partia...
متن کاملAppearance-Based 3-D Face Recognition from Video
In this work we present an appearance-based 3-D Face Recognition approach that is able to recognize faces in video sequences, independent from face pose. For this we combine eigen light-fields with probabilistic propagation over time for evidence integration. Eigen light-fields allow us to build an appearance based 3-D model of an object; probabilistic methods for evidence integration are attra...
متن کاملIntegrating monolithic and free-parts representations for improved face verification in the presence of pose mismatch
Face images, varying under pose, are dramatically different in their “pixel” appearance even if they stem from the same subject. Our work concentrates specifically on the task of verifying faces when the gallery set stems from frontal face images, with the probe set stemming from a number of alternate poses (i.e. pose mismatch). An argument is put forward for attempting to recognize faces throu...
متن کاملEigen space approach for a pose detection with range images
Abstract An application of the Pose-detection using Range-image usually uses characteristic matching of the geometrical model but this method has two problems. 1 Picking for characteristic from Range-images is difficult. 2 It is difficult to make geometrical model for a complicated shape. Recently Parametric Eigen-space Method is proposed to Pose-detect approaching such as Eigen-face. This meth...
متن کامل