Depthwise Spatio-Temporal STFT Convolutional Neural Networks for Human Action Recognition

TL;DR

This paper introduces spatio-temporal STFT blocks as a computationally efficient alternative to traditional 3D CNN layers, improving feature learning for human action recognition while reducing parameters and computational costs.

Contribution

The paper proposes a novel STFT block for 3D CNNs that captures Fourier information with fewer parameters and better feature learning capabilities, outperforming conventional methods.

Findings

STFT-based 3D CNNs use 3.5 to 4.5 times fewer parameters.

They achieve 1.5 to 1.8 times less computational cost.

Performance is on par or better than state-of-the-art methods on multiple datasets.

Abstract

Conventional 3D convolutional neural networks (CNNs) are computationally expensive, memory intensive, prone to overfitting, and most importantly, there is a need to improve their feature learning capabilities. To address these issues, we propose spatio-temporal short term Fourier transform (STFT) blocks, a new class of convolutional blocks that can serve as an alternative to the 3D convolutional layer and its variants in 3D CNNs. An STFT block consists of non-trainable convolution layers that capture spatially and/or temporally local Fourier information using a STFT kernel at multiple low frequency points, followed by a set of trainable linear weights for learning channel correlations. The STFT blocks significantly reduce the space-time complexity in 3D CNNs. In general, they use 3.5 to 4.5 times less parameters and 1.5 to 1.8 times less computational costs when compared to the state-of-the-art methods. Furthermore, their feature learning capabilities are significantly better than the conventional 3D convolutional layer and its variants. Our extensive evaluation on seven action recognition datasets, including Something-something v1 and v2, Jester, Diving-48, Kinetics-400, UCF 101, and HMDB 51, demonstrate that STFT blocks based 3D CNNs achieve on par or even better performance compared to the state-of-the-art methods.