MVFNet: Multi-View Fusion Network for Efficient Video Recognition
This work addresses the challenge of achieving both high accuracy and computational efficiency in video action recognition, which is a significant problem for researchers and practitioners working with video data.
The paper introduces MVFNet, a multi-view fusion network for video action recognition that models video from three different planes (Height-Width, Height-Time, and Width-Time) to capture dynamics efficiently. MVFNet achieves state-of-the-art performance with the complexity of 2D CNNs on popular benchmarks like Something-Something V1 & V2, Kinetics, UCF-101, and HMDB-51.
Conventionally, spatiotemporal modeling network and its complexity are the two most concentrated research topics in video action recognition. Existing state-of-the-art methods have achieved excellent accuracy regardless of the complexity meanwhile efficient spatiotemporal modeling solutions are slightly inferior in performance. In this paper, we attempt to acquire both efficiency and effectiveness simultaneously. First of all, besides traditionally treating H x W x T video frames as space-time signal (viewing from the Height-Width spatial plane), we propose to also model video from the other two Height-Time and Width-Time planes, to capture the dynamics of video thoroughly. Secondly, our model is designed based on 2D CNN backbones and model complexity is well kept in mind by design. Specifically, we introduce a novel multi-view fusion (MVF) module to exploit video dynamics using separable convolution for efficiency. It is a plug-and-play module and can be inserted into off-the-shelf 2D CNNs to form a simple yet effective model called MVFNet. Moreover, MVFNet can be thought of as a generalized video modeling framework and it can specialize to be existing methods such as C2D, SlowOnly, and TSM under different settings. Extensive experiments are conducted on popular benchmarks (i.e., Something-Something V1 & V2, Kinetics, UCF-101, and HMDB-51) to show its superiority. The proposed MVFNet can achieve state-of-the-art performance with 2D CNN's complexity.