A Self-Adaptive Gallery Construction Method for Open-World Person Re-Identification.

Person re-identification, or simply re-id, is the task of identifying again a person who has been seen in the past by a perception system. Multiple robotic applications, such as tracking or navigate-and-seek, use re-identification systems to perform their tasks. To solve the re-id problem, a common practice consists in using a gallery with relevant information about the people already observed. The construction of this gallery is a costly process, typically performed offline and only once because of the problems associated with labeling and storing new data as they arrive in the system. The resulting galleries from this process are static and do not acquire new knowledge from the scene, which is a limitation of the current re-id systems to work for open-world applications. Different from previous work, we overcome this limitation by presenting an unsupervised approach to automatically identify new people and incrementally build a gallery for open-world re-id that adapts prior knowledge with new information on a continuous basis. Our approach performs a comparison between the current person models and new unlabeled data to dynamically expand the gallery with new identities. We process the incoming information to maintain a small representative model of each person by exploiting concepts of information theory. The uncertainty and diversity of the new samples are analyzed to define which ones should be incorporated into the gallery. Experimental evaluation in challenging benchmarks includes an ablation study of the proposed framework, the assessment of different data selection algorithms that demonstrate the benefits of our approach, and a comparative analysis of the obtained results with other unsupervised and semi-supervised re-id methods.

Human-computer interaction based on hand gestures using RGB-D sensors.

In this paper we present a new method for hand gesture recognition based on an RGB-D sensor. The proposed approach takes advantage of depth information to cope with the most common problems of traditional video-based hand segmentation methods: cluttered backgrounds and occlusions. The algorithm also uses colour and semantic information to accurately identify any number of hands present in the image. Ten different static hand gestures are recognised, including all different combinations of spread fingers. Additionally, movements of an open hand are followed and 6 dynamic gestures are identified. The main advantage of our approach is the freedom of the user's hands to be at any position of the image without the need of wearing any specific clothing or additional devices. Besides, the whole method can be executed without any initial training or calibration. Experiments carried out with different users and in different environments prove the accuracy and robustness of the method which, additionally, can be run in real-time.