Learning Gating ConvNet for Two-Stream based Methods in Action Recognition

TL;DR

This paper introduces an end-to-end trainable gating ConvNet for two-stream action recognition that adaptively fuses spatial and temporal features, improving accuracy over fixed-weight methods.

Contribution

It proposes a novel gating ConvNet based on MoE theory with multi-task learning to enhance fusion adaptability and reduce overfitting in two-stream action recognition models.

Findings

Achieved 94.5% accuracy on UCF101 dataset.

Outperforms fixed-weight fusion methods.

Demonstrates improved adaptability and robustness.

Abstract

For the two-stream style methods in action recognition, fusing the two streams' predictions is always by the weighted averaging scheme. This fusion method with fixed weights lacks of pertinence to different action videos and always needs trial and error on the validation set. In order to enhance the adaptability of two-stream ConvNets and improve its performance, an end-to-end trainable gated fusion method, namely gating ConvNet, for the two-stream ConvNets is proposed in this paper based on the MoE (Mixture of Experts) theory. The gating ConvNet takes the combination of feature maps from the same layer of the spatial and the temporal nets as input and adopts ReLU (Rectified Linear Unit) as the gating output activation function. To reduce the over-fitting of gating ConvNet caused by the redundancy of parameters, a new multi-task learning method is designed, which jointly learns the gating fusion weights for the two streams and learns the gating ConvNet for action classification. With our gated fusion method and multi-task learning approach, a high accuracy of 94.5% is achieved on the dataset UCF101.