A Spatiotemporal Oriented Energy Network for Dynamic Texture Recognition

TL;DR

This paper introduces a non-learning, hierarchical spatiotemporal orientation network that analytically captures dynamic textures, achieving state-of-the-art results in texture recognition without relying on training.

Contribution

It proposes a novel, analytically designed hierarchical spatiotemporal network with recurrent feedback and cross-channel pooling, advancing dynamic texture recognition methods.

Findings

Achieves state-of-the-art performance on standard datasets.

Provides a theoretically motivated, interpretable analysis of spatiotemporal features.

Demonstrates effectiveness without learning, reducing heuristic design choices.

Abstract

This paper presents a novel hierarchical spatiotemporal orientation representation for spacetime image analysis. It is designed to combine the benefits of the multilayer architecture of ConvNets and a more controlled approach to spacetime analysis. A distinguishing aspect of the approach is that unlike most contemporary convolutional networks no learning is involved; rather, all design decisions are specified analytically with theoretical motivations. This approach makes it possible to understand what information is being extracted at each stage and layer of processing as well as to minimize heuristic choices in design. Another key aspect of the network is its recurrent nature, whereby the output of each layer of processing feeds back to the input. To keep the network size manageable across layers, a novel cross-channel feature pooling is proposed. The multilayer architecture that results systematically reveals hierarchical image structure in terms of multiscale, multiorientation properties of visual spacetime. To illustrate its utility, the network has been applied to the task of dynamic texture recognition. Empirical evaluation on multiple standard datasets shows that it sets a new state-of-the-art.