Topology dictionary with Markov model for 3D video content-based skimming and description

Abstract

This paper presents a novel approach to skim and describe 3D videos. 3D video is an imaging technology which consists in a stream of 3D models in motion captured by a synchronized set of video cameras. Each frame is composed of one or several 3D models, and therefore the acquisition of long sequences at video rate requires massive storage devices. In order to reduce the storage cost while keeping relevant information, we propose to encode 3D video sequences using a topology-based shape descriptor dictionary. This dictionary is either generated from a set of extracted patterns or learned from training input sequences with semantic annotations. It relies on an unsupervised 3D shapebased clustering of the dataset by Reeb graphs, and features a Markov network to characterize topological changes. The approach allows content-based compression and skimming with accurate recovery of sequences and can handle complex topological changes. Redundancies are detected and skipped based on a probabilistic discrimination process. Semantic description of video sequences is then automatically performed. In addition, forthcoming frame encoding is achieved using a multiresolution matching scheme and allows action recognition in 3D. Our experiments were performed on complex 3D video sequences. We demonstrate the robustness and accuracy of the 3D video skimming with dramatic low bitrate coding and high compression ratio.

Topology Dictionary for 3D Video Understanding

“…To our knowledge, although skimming techniques are successful for 2D videos [35], [36], the extension to 3D videos with automatic segmentation, encoding, and reconstruction has not been treated yet. Preliminary versions of this paper appeared in [15] and [37].…”

“…As in [37], a 3D shape descriptor is used to identify similar poses, and a graph is built to represent potential transitions. However, in [38], a 3D histogram is used as 3D shape descriptor.…”