Retain, Blend, and Exchange: A Quality-aware Spatial-Stereo Fusion Approach for Event Stream Recognition
This work addresses performance limitations in event stream recognition for applications like surveillance or robotics, though it appears incremental as it builds on existing dual-stream and fusion methods.
The paper tackles the problem of limited performance in event stream-based pattern recognition due to monotonous modality expressions and sub-optimal fusion by proposing EFV++, a dual-stream framework that models event images and voxels separately using Transformer and GNN, then fuses features with quality-aware retention, blending, and exchange. It achieves state-of-the-art performance, such as 90.51% accuracy on the Bullying10k dataset, exceeding the second place by 2.21%.
Existing event stream-based pattern recognition models usually represent the event stream as the point cloud, voxel, image, etc., and design various deep neural networks to learn their features. Although considerable results can be achieved in simple cases, however, the model performance may be limited by monotonous modality expressions, sub-optimal fusion, and readout mechanisms. In this paper, we propose a novel dual-stream framework for event stream-based pattern recognition via differentiated fusion, termed EFV++. It models two common event representations simultaneously, i.e., event images and event voxels. The spatial and three-dimensional stereo information can be learned separately by utilizing Transformer and Graph Neural Network (GNN). We believe the features of each representation still contain both efficient and redundant features and a sub-optimal solution may be obtained if we directly fuse them without differentiation. Thus, we divide each feature into three levels and retain high-quality features, blend medium-quality features, and exchange low-quality features. The enhanced dual features will be fed into the fusion Transformer together with bottleneck features. In addition, we introduce a novel hybrid interaction readout mechanism to enhance the diversity of features as final representations. Extensive experiments demonstrate that our proposed framework achieves state-of-the-art performance on multiple widely used event stream-based classification datasets. Specifically, we achieve new state-of-the-art performance on the Bullying10k dataset, i.e., $90.51\%$, which exceeds the second place by $+2.21\%$. The source code of this paper has been released on \url{https://github.com/Event-AHU/EFV_event_classification/tree/EFVpp}.