Multiple Face Tracking with Appearance Modes and Reasoning
نویسندگان
چکیده
Multiple face tracking plays a key role in applications related to security-surveillance, human-computer interactions, video indexing etc. Existing literature in face tracking has mainly focused on facial features from a detection/recognition viewpoint. On the other hand, we believe that reasoning with detected/tracked face regions has a strong role in multiple face tracking. We propose a reasoning scheme that binds face localization (motion and mean-shift) and detection (Ada-Boost with Haar features) for tracking multiple faces in image sequences. The reasoning procedure identifies the cases of face isolation (unoccluded), grouping/occlusions, detection/tracking failure, entry/exit and reappearance of faces. Instantiation of these cases are used as cues in selective update of the facial features. Additionally, we maintain a Normalized Face Cluster Set (NFCS) to capture the appearance modes for varying facial poses. These cluster sets are further used in discriminating new faces from the existing ones while restoring the tracks of the later. Experimental validation on four video sequences has shown significant tracking performance under occlusions.
منابع مشابه
Online multiple people tracking-by-detection in crowded scenes
Multiple people detection and tracking is a challenging task in real-world crowded scenes. In this paper, we have presented an online multiple people tracking-by-detection approach with a single camera. We have detected objects with deformable part models and a visual background extractor. In the tracking phase we have used a combination of support vector machine (SVM) person-specific classifie...
متن کاملRobust Face Tracking using Multiple Appearance Models and Graph Relational Learning
This paper addresses the problem of appearance matching across different challenges while doing visual face tracking in real-world scenarios. In this paper, FaceTrack is proposed that utilizes multiple appearance models with its long-term and short-term appearance memory for efficient face tracking. It demonstrates robustness to deformation, in-plane and out-of-plane rotation, scale, distractor...
متن کاملTheory of evidence for face detection and tracking
This paper deals with face detection and tracking by computer vision for multimedia applications. Contrary to current techniques that are based on huge learning databases and complex algorithms to get generic face models (e.g. active appearance models), the proposed method handles simple contextual knowledge representative of the application background thanks to a quick supervised initializatio...
متن کاملRobust 3D Face Tracking on Unknown Users with Dynamical Active Models
The Active Appearance Models [1] and the derived Active Models (AM) [4] allow to robustly track the face of a single user that was previously learnt, but works poorly with multiple or unknown users. Our research aims at improving the tracking robustness by learning from video databases. In this paper, we study the relation between the face texture and the parameter gradient matrix, and propose ...
متن کاملMultiple Target Tracking With a 2-D Radar Using the JPDAF Algorithm and Combined Motion Model
Multiple target tracking (MTT) is taken into account as one of the most important topics in tracking targets with radars. In this paper, the MTT problem is used for estimating the position of multiple targets when a 2-D radar is employed to gather measurements. To do so, the Joint Probabilistic Data Association Filter (JPDAF) approach is applied to tracking the position of multiple targets. To ...
متن کامل