Local Invariant Feature Tracks for High-Level Video Feature Extraction

Abstract

This paper builds upon previous work on local interest point detection and description to propose the extraction and representation of novel Local Invariant Feature Tracks (LIFT). These features compactly capture not only the spatial attributes of 2D local regions, as in SIFT and related techniques, but also their long-term trajectories in time. This and other desirable properties of LIFT allow the generation of Bags-of-Spatiotemporal-Words models that facilitate capturing the dynamics of video content, which is necessary for detecting high-level video features that by definition have a strong temporal dimension. Preliminary experimental evaluation and comparison of the proposed approach reveals promising results.

Trajectories based descriptor for dynamic events annotation

“…A trajectory or track is a set of local 2D features found in successive frames of a video that are constraints by space-time and visual appearance continuity [17]. Tracks detection follows the same scheme amongst the existing works.…”

“…[23,16] use the trajectories motion transition to characterize their track descriptors. Multiple Haar filters are used by [17] to capture the multiscale aspect of the trajectory motions. An aggregating scheme is then used to combine the different tracks descriptors in one global representation.…”

“…BoW based on trajectory features [23,15,19,16,17] have been introduced to capture the motion patterns in their duration. A trajectory or track is a set of local 2D features found in successive frames of a video that are constraints by space-time and visual appearance continuity [17].…”

“…In this work, we focus on the recognition of dynamics events (running, walking) in unconstrained videos. Feature trajectories have been shown to be an efficient video representation [23,16,17]. Trajectories are extracted by tracking interest points over several video frames to capture both motion and appearance information.…”

A Joint Content-Event Model for Event-Centric Multimedia Indexing

“…Many works have appeared on combining the aforementioned low-level features with machine learning algorithms in order to achieve the association of content with concepts such as "person", "outdoors", etc., or different actions [15], [16]. The content segments can then be retrieved with the use of the detected concepts [13].…”