In recent years, event-based deep learning has garnered widespread attention among researchers. Compared to conventional visual data, event data exhibits unique advantages in terms of temporal resolution and adaptability to dynamic lighting conditions. However, the extreme sparsity of event streams over prolonged periods limits effective feature extraction, while existing data augmentation methods fail to adequately exploit latent spatiotemporal patterns. To address this issue, three data augmentation strategies were proposed: temporal chunking, random beginning time, and geometric dimension-wise random drop. These methods comprehensively augment event data along both temporal and geometric dimensions, thereby mitigating the limitations of sparse event data in latent feature extraction. Experimental results on three event-based datasets including DVS-Gesture, DVS128-Gait-Day, and DVS-CIFAR10 demonstrate that the proposed methods enable models to achieve superior recognition performance under short input time windows, significantly improving the utilization efficiency of event data. Compared to baseline methods, the classification performance is enhanced by 17.62%, 18.5%, and 9.7%, respectively.
LI Hao-fei, ZHANG Xiao-song, YANG Wei, WANG Yi-peng, YAO Jia-lin, WANG Ding-heng.
Multi-dimensional Augmentation Methods for Event Data in Dynamic Vision Recognition[J].
Science Technology and Engineering, 2026, 26(14): 6045-6055 DOI:10.12404/j.issn.1671-1815.2503912
LuoX, YaoM, ChouY, et al. Integer-valued training and spike-driven inference spiking neural network for high-performance and energy-efficient object detection[C]//European Conference on Computer Vision. Cham: Springer, 2024: 253-272.
[5]
ZhangJ, DongB, ZhangH, et al. Spiking transformers for event-based single object tracking[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. New Orleans: IEEE, 2022: 8801-8810.
[6]
MaassW. Networks of spiking neurons: the third generation of neural network models[J]. Neural Networks, 1997, 10(9): 1659-1671.
[7]
GuoY, ChenY, ZhangL, et al. Im-loss: information maximization loss for spiking neural networks[J]. Advances in Neural Information Processing Systems, 2022, 35: 156-166.
ZhangChaoqun, YiYunheng, ZhouWenjuan, et al. Small rock samples classification based on deep learning and data enhancement technologies[J]. Science Technology and Engineering, 2022, 22(33): 14786-14794.
TangHaoyao, CuiXin, ZhangYiwei, et al. Network traffic classification method improved based on data augmentation and CNN-Optuna-Attention[J]. Science Technology and Engineering, 2025, 25(15): 6419-6430.
SunJiajia, LiChengli, ChangDexian, et al. Data augmentation method for intrusion detection imbalance problem using generative adversarial networks[J]. Science Technology and Engineering, 2022, 22(18): 7965-7971.
LinZhipeng, ZengLibo, WuQiongshui. Image data augmentation of cervical cells based on generative adversarial networks[J]. Science Technology and Engineering, 2020, 20(28): 11672-11677.
[18]
LiY, KimY, ParkH, et al. Neuromorphic data augmentation for training spiking neural networks[C]//European Conference on Computer Vision. Cham: Springer, 2022: 631-649.
[19]
BendigK, SchusterR, StrickerD. Shapeaug: occlusion augmentation for event camera data[J]. arXiv Preprint, 2024, arXiv: 2401.02274.
[20]
ShenG, ZhaoD, ZengY. Eventmix: an efficient data augmentation strategy for event-based learning[J]. Information Sciences, 2023, 644: 119170.
[21]
YunS, HanD, OhS J, et al. Cutmix: regularization strategy to train strong classifiers with localizable features[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision. Seoul: IEEE, 2019: 6023-6032.
[22]
DeVriesT, TaylorG W. Improved regularization of convolutional neural networks with cutout[J]. arXiv Preprint, arXiv: 1708.04552, 2017.
[23]
GuF, SngW, HuX, et al. Eventdrop: data augmentation for event-based learning[J]. arXiv Preprint, arXiv: 2106.05836, 2021.
[24]
SunM, ZhangD, GeZ, et al. Eventrpg: event data augmentation with relevance propagation guidance[J]. arXiv Preprint, arXiv: 2403. 09274, 2024.
[25]
RebecqH, HorstschaeferT, ScaramuzzaD. Real-time visual-inertial odometry for event cameras using keyframe-based nonlinear optimization[C]//British Machine Vision Conference. London: British Machine Vision Association, 2017: 1-8.
[26]
FangW, YuZ, ChenY, et al. Incorporating learnable membrane time constant to enhance learning of spiking neural networks[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision. Virtual: IEEE, 2021: 2661-2671.
[27]
DengL, WuY, HuX, et al. Rethinking the performance comparison between SNNs and ANNs[J]. Neural Networks, 2020, 121: 294-307.
[28]
HeW, WuY J, DengL, et al. Comparing SNNs and RNNs on neuromorphic vision datasets: similarities and differences[J]. Neural Networks, 2020, 132: 108-120.
[29]
YaoM, GaoH, ZhaoG, et al. Temporal-wise attention spiking neural networks for event streams classification[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision. Virtual: IEEE, 2021: 10221-10230.
[30]
AmirA, TabaB, BergD, et al. A low power, fully event-based gesture recognition system[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Honolulu: IEEE, 2017: 7243-7252.
[31]
WangY, ZhangX, ShenY, et al. Event-stream representation for human gaits identification using deep neural networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021, 44(7): 3436-3449.
[32]
LiH, LiuH, JiX, et al. CIFAR10-DVS: an event-stream dataset for object classification[J]. Frontiers in Neuroscience, 2017, 11: 1-10.
[33]
ZhengH, WuY, DengL, et al. Going deeper with directly-trained larger spiking neural networks[C]//Proceedings of the AAAI Conference on Artificial Intelligence. Virtual: AAAI, 2021: 11062-11070.
[34]
WangY, DuB, ShenY, et al. Ev-Gait: event-based robust gait recognition using dynamic vision sensors[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach: IEEE, 2019: 6358-6367.
[35]
YaoM, ZhangH, ZhaoG, et al. Sparser spiking activity can be better: feature refine-and-mask spiking neural network for event-based visual recognition[J]. Neural Networks, 2023, 166: 410-423.
[36]
WuY, DengL, LiG, et al. Spatio-temporal backpropagation for training high-performance spiking neural networks[J]. Frontiers in Neuroscience, 2018, 12: 1-12.