Optimized Culprit Identification Using Mobilenet and Attention Mechanisms

TL;DR

This paper presents a lightweight MobileNet-based deep learning framework with attention mechanisms for accurate, real-time culprit identification in surveillance, demonstrating high accuracy and efficiency on benchmark face datasets.

Contribution

The novel integration of attention mechanisms into MobileNet enhances feature focus, improving identification accuracy while maintaining computational efficiency for real-time use.

Findings

Achieved 97.8% classification accuracy on benchmark datasets.

Outperformed baseline CNN, ResNet, and standard MobileNet models.

Maintained low inference time suitable for real-time applications.

Abstract

Automated culprit identification in surveillance systems is a critical task that requires high accuracy along with computational efficiency for real-time deployment. In this paper, an optimized deep learning framework is proposed using a lightweight MobileNet architecture integrated with channel and spatial attention mechanisms. The proposed model enhances feature representation by selectively focusing on the most discriminative regions while suppressing irrelevant background information, thereby improving identification performance. The framework incorporates efficient preprocessing, attention based feature refinement, and a robust classification strategy optimized using the Adam Optimizer. Experiments were conducted on benchmark face recognition datasets, including Labelled Faces in the Wild (LFW), CASIA-WebFace, and a subset of VGGFace2, under realistic conditions with variations in illumination, pose, and occlusion. The results demonstrate that the proposed model achieves a high classification accuracy of 97.8%, outperforming conventional models such as baseline CNN, ResNet, and standard MobileNet. The confusion matrix analysis indicates strong class-wise discrimination with minimal misclassification, while ROC-AUC evaluation confirms robust performance across all classes. Additionally, the proposed approach maintains low computational complexity and reduced inference time, making it suitable for real-time surveillance and edge-based applications.