Scholarlink

This study investigated three bio-inspired optimization techniques, namely Genetic Algorithm (GA), Particle Swarm Optimization (PSO) and Differential Evolution (DE), in order to optimize key hyperparameters (learning rate and epochs) of the 3D Convolutional Neural Network (Simple3DCNN) for video anomaly detection, based on the testing of the RWF-2000 dataset. The overall findings showed that PSO achieved superior F1 scores, which means that it was able to achieve a better balance of exploration and exploitation of the search space. In conclusion, a combination of lightweight 3D CNN and intelligent optimization algorithms is a powerful framework toward establishment of a real-time video anomaly detection system.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Comparative Study of Bio-inspired Algorithms for Video Anomaly Detection Using 3D CNN-Based Deep Learning

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • J Monisha,
  • Karthika Subbaraj

摘要

This study investigated three bio-inspired optimization techniques, namely Genetic Algorithm (GA), Particle Swarm Optimization (PSO) and Differential Evolution (DE), in order to optimize key hyperparameters (learning rate and epochs) of the 3D Convolutional Neural Network (Simple3DCNN) for video anomaly detection, based on the testing of the RWF-2000 dataset. The overall findings showed that PSO achieved superior F1 scores, which means that it was able to achieve a better balance of exploration and exploitation of the search space. In conclusion, a combination of lightweight 3D CNN and intelligent optimization algorithms is a powerful framework toward establishment of a real-time video anomaly detection system.