基于小样本学习的老年人异常行为识别模型研究
Research on Recognition Model of Abnormal Behavior of the Elderly Based on Few-Shot Learning
DOI: 10.12677/CSA.2023.133045, PDF,   
作者: 王聪敏：天津商业大学理学院，天津；周艳聪：天津商业大学信息工程学院，天津
关键词: 异常行为识别；小样本学习；深度学习；Abnormal Behavior Recognition； Few-Shot Learning； Deep Learning
摘要: 当今社会老龄化程度逐步加重，老年人的看护问题日益凸显。基于视频识别的老年人异常行为识别旨在助力老年人看护问题。但该领域相关公开的老年人异常行为数据稀缺。针对这种情况，本文收集了老年人异常行为数据，并结合小样本学习设计了老年人异常行为识别模型(RAAE, Recognition model of Abnormal Action of the Elderly)。该网络通过patch级富集模块获取patch级空间信息，后接通道注意力模块获取帧间时间语义信息，最后通过CrossTransformer进行时间建模。实验表明该网络和以往模型相比，对老年人异常行为的识别准确率得到了4.7%的提升。
Abstract: Nowadays, the aging of society is gradually increasing, and the care of the elderly is increasingly prominent. The recognition of abnormal behaviors of the elderly based on video recognition aims to help the elderly care problems. However, relevant public data on abnormal behavior of the elderly in this field are scarce. In response to this situation, this paper collected the data of abnormal behavior of the elderly, and designed a recognition model of abnormal behavior of the elderly (RAAE, Recognition model of Abnormal Action of the Elderly) based on few-shot learning. The network obtains the patch-level spatial information through the patch-level enrichment module, and then ob-tains the inter-frame temporal semantic information through the channel attention module. Finally, time modeling is performed through Cross Transformer. The experiment shows that compared with previous models, the recognition accuracy of the network for abnormal behavior of the elderly has been improved by 4.7%.
文章引用：王聪敏, 周艳聪. 基于小样本学习的老年人异常行为识别模型研究[J]. 计算机科学与应用, 2023, 13(3): 465-471. https://doi.org/10.12677/CSA.2023.133045

参考文献

[1] Lentzas, A. and Vrakas, D. (2020) Non-Intrusive Human Activity Recognition and Abnormal Behavior Detection on El-derly People: A Review. Artificial Intelligence Review, 53, 1975-2021. [Google Scholar] [CrossRef] 
[2] 于乃功, 柏德国. 基于姿态估计的实时跌倒检测算法[J]. 控制与决策, 2020, 35(11): 2761-2766.
[3] 黄加成. 中国老龄化社会独居老人研究进展[J]. 中国老年学杂志, 2015, 35(23): 6954-6956.
[4] Mubashir, M., Shao, L. and Seed, L. (2013) A Survey on Fall Detection: Principles and Ap-proaches. Neurocomputing, 100, 144-152. [Google Scholar] [CrossRef] 
[5] 李饶. 老年人群居家异常行为检测技术研究[D]: [硕士学位论文]. 武汉: 华中科技大学, 2021.
[6] 王宇, 周迎, 王灵灵, 等. 室内环境下老年人异常行为检测[J]. 土木工程与管理学报, 2022, 39(4): 145-152.
[7] 李一凡. 基于Wi-Fi的非接触式老年人异常行为识别研究[D]: [硕士学位论文]. 兰州: 西北民族大学, 2022.
[8] 彭小英. 室内环境下老年人异常行为检测系统设计与研究[D]: [硕士学位论文]. 武汉: 华中科技大学, 2020.
[9] 温佳昆. 基于机器学习的老年人异常行为检测方法研究[D]: [硕士学位论文]. 沈阳: 沈阳理工大学, 2021.
[10] 孟祥泽, 胡啸峰, 沈兵. 社区老年人空间行为轨迹异常分析方法[J]. 科学技术与工程, 2021, 21(9): 3676-3681.
[11] 张晓平, 纪佳慧, 王力, 等. 基于视频的人体异常行为识别与检测方法综述[J]. 控制与决策, 2022, 37(1): 14-27.
[12] 杨鹏, 李潇婧, 张慧君, 等. 智能监控系统下的老年人异常行为研究进展[J]. 中国康复医学杂志, 2019, 34(1): 112-115.
[13] 赵凯琳, 靳小龙, 王元卓. 小样本学习研究综述[J]. 软件学报, 2021, 32(2): 349-369.
[14] 韦世红, 刘红梅, 唐宏, 朱龙娇. 多级度量网络的小样本学习[J]. 计算机工程与应用, 2023, 59(2): 94-101.
[15] http://fenix.univ.rzeszow.pl/~mkepski/ds/uf.html
[16] http://falldataset.com
[17] http://mha.cs.umn.edu/projevents.shtml#crowd
[18] Luo, W.X., Liu, L. and Gao, S.H. (2017) A Revisit of Sparse Coding Based Anomaly De-tection in Stacked RNN Framework. 2017 IEEE International Conference on Computer Vision (ICCV), Venice, 22-29 October 2017, 341-349. [Google Scholar] [CrossRef] 
[19] Lv, H., Zhou, C.W., Xu, C.Y., et al. (2021) Localizing Anomalies from Weakly-Labeled Videos. IEEE Transactions on Image Processing, 30, 4505-4515.
[20] Perrett, T., Masullo, A., Burghardt, T., Mirmehdi, M. and Damen, D. (2021) Temporal-Relational CrossTransformers for Few-Shot Action Recognition. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, 20-25 June 2021, 475-484. [Google Scholar] [CrossRef] 
[21] He, K.M., Zhang, X.Y., Ren, S.Q. and Sun, J. (2016) Deep Residual Learning for Image Recognition. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, 27-30 June 2016, 770-778.
[22] Doersch, C., Gupta, A. and Zisserman, A. (2020) CrossTransformers: Spatially-Aware Few-Shot Transfer. 34th Conference on Neural Information Processing Systems (NeurIPS 2020), Vancouver, 6-12 December 2020, 1-3+5.
[23] Woo, S., Park, J., Lee, J.-Y. and Kweon, I.S. (2018) CBAM: Convolutional Block Attention Module. Computer Vision—ECCV 2018, Munich, 8-14 September 2018, 3-19.
[24] Liu, J., Shahroudy, A., Perez, M., Wang, G., Duan, L.-Y. and Kot, A.C. (2019) NTU RGB+D 120: A Large-Scale Benchmark for 3D Human Activity Understanding. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 42, 2684-2701. [Google Scholar] [CrossRef]