Multi-view Discriminative Fusion on Canonical Correlation Analysis in Event Classification

Video event carries very rich and complex semantics, and video event analysis is intrinsically multi-view learning problem. In this paper, by revealing the relationship between canonical correlation analysis (CCA) and linear discriminant analysis (LDA), we propose a new convenient multi-view learning architecture, i.e., multi-view discriminative fusion on CCA (MvDF-CCA). Instead of pair-wise relation consideration among different views, MvDF-CCA separately aligns each view space to the target one comprised by labeled indicators, which is easy expanding to additional view data. MvDF-CCA leverages the discriminative ability in LDA and the strengths in CCA, and maps all view spaces to the same target labeled space, which can be concurrent execution on large scale video categorization. Competitive results are reported on the well-known UCF101 and Columbia Consumer Video (CCV) benchmarks.

I. Introduction

Human can understand the world in a single glance, in which the recognition ability does step from different aspect perceiving. To recognize an object, people firstly notice the shape and contour profile of the object, then care about its appearance and texture information [1]. Both shape and appearance help people to construct the concept of this object in brain. While for scene distinguishing, it includes not only shape and appearance information, but also the object relative locations in the whole picture and so on [2]. For video event analysis, it needs much more clues whose comprehensive effects figure out the event concept. How to combine different clues becomes a big problem in video/image processing, which is also important to data fusing in multi-sensor network.