Action Recognition based on Subdivision-Fusion Model

TL;DR

This paper introduces a Subdivision-Fusion Model (SFM) for human action recognition that improves accuracy by clustering samples into subcategories to reduce overfitting and enhance classification.

Contribution

The novel SFM approach clusters samples into subcategories to address feature overlap and overfitting, with methods for determining cluster numbers and demonstrated superior performance.

Findings

Achieved 79.4% accuracy on Hollywood2, surpassing 64.3% state-of-the-art.

Improved YouTube Action dataset accuracy from 75.8% to 82.5%.

Significant improvements on KTH and UCF50 datasets.

Abstract

This paper proposes a novel Subdivision-Fusion Model (SFM) to recognize human actions. In most action recognition tasks, overlapping feature distribution is a common problem leading to overfitting. In the subdivision stage of the proposed SFM, samples in each category are clustered. Then, such samples are grouped into multiple more concentrated subcategories. Boundaries for the subcategories are easier to find and as consequence overfitting is avoided. In the subsequent fusion stage, the multi-subcategories classification results are converted back to the original category recognition problem. Two methods to determine the number of clusters are provided. The proposed model has been thoroughly tested with four popular datasets. In the Hollywood2 dataset, an accuracy of 79.4% is achieved, outperforming the state-of-the-art accuracy of 64.3%. The performance on the YouTube Action dataset has been improved from 75.8% to 82.5%, while considerably improvements are also observed on the KTH and UCF50 datasets.