2 resultados para Visual surveillance, Human activity recognition, Video annotation
em Digital Peer Publishing
Resumo:
This paper presents an empirical study of affine invariant feature detectors to perform matching on video sequences of people with non-rigid surface deformation. Recent advances in feature detection and wide baseline matching have focused on static scenes. Video frames of human movement capture highly non-rigid deformation such as loose hair, cloth creases, skin stretching and free flowing clothing. This study evaluates the performance of six widely used feature detectors for sparse temporal correspondence on single view and multiple view video sequences. Quantitative evaluation is performed of both the number of features detected and their temporal matching against and without ground truth correspondence. Recall-accuracy analysis of feature matching is reported for temporal correspondence on single view and multiple view sequences of people with variation in clothing and movement. This analysis identifies that existing feature detection and matching algorithms are unreliable for fast movement with common clothing.
Resumo:
Though migration is an age-old feature of human activity, driven by various circumstances, its current place in the midst of global dynamics and the phenomenon of globalization is becoming increasingly critical. International immigration and its regulation have been largely shaped by the policies in the receiving countries, often determining preferences for nationality cohorts and work skills to satisfy their labor and human capital requirements. When immigration has been necessitated by political strife, host countries have displayed immense magnanimity as well. However, the growing realization of resource limitations and the strange quirks of cultural pluralism are in turn creating waves of dissonance. Literature and the media are now replete with an in depth look into the immigration debate in various nations of the world in trying to seek new directions and satisfactory solutions.