Eigen Evolution Pooling for Human Action Recognition

TL;DR

Eigen Evolution Pooling is a novel method for aggregating sequential feature vectors that effectively captures temporal dynamics, leading to improved human action recognition performance on benchmark datasets.

Contribution

The paper introduces Eigen Evolution Pooling, a new general pooling technique that preserves temporal information in feature sequences for action recognition.

Findings

Outperforms average, max, and rank pooling methods.

Achieves state-of-the-art results on UCF101 and Hollywood2 datasets.

Effectively encodes temporal evolution of features.

Abstract

We introduce Eigen Evolution Pooling, an efficient method to aggregate a sequence of feature vectors. Eigen evolution pooling is designed to produce compact feature representations for a sequence of feature vectors, while maximally preserving as much information about the sequence as possible, especially the temporal evolution of the features over time. Eigen evolution pooling is a general pooling method that can be applied to any sequence of feature vectors, from low-level RGB values to high-level Convolutional Neural Network (CNN) feature vectors. We show that eigen evolution pooling is more effective than average, max, and rank pooling for encoding the dynamics of human actions in video. We demonstrate the power of eigen evolution pooling on UCF101 and Hollywood2 datasets, two human action recognition benchmarks, and achieve state-of-the-art performance.