12 February 2021 in IEEE Access
Security and safety is a big concern for today's modern world. For a country to be economically strong, it must ensure a safe and secure environment for investors and tourists. Having said that, Closed Circuit Television (CCTV) cameras are being used for surveillance and to monitor activities i.e. robberies but these cameras still require human supervision and intervention. We need a system that can automatically detect these illegal activities. Despite state-of-the-art deep learning algorithms, fast processing hardware, and advanced CCTV cameras, weapon detection in real-time is still a serious challenge. Observing angle differences, occlusions by the carrier of the firearm and persons around it further enhances the difficulty of the challenge. This work focuses on providing a secure place using CCTV footage as a source to detect harmful weapons by applying the state of the art open-source deep learning algorithms. We have implemented binary classification assuming pistol class as the reference class and relevant confusion objects inclusion concept is introduced to reduce false positives and false negatives. No standard dataset was available for real-time scenario so we made our own dataset by making weapon photos from our own camera, manually collected images from internet, extracted data from YouTube CCTV videos, through GitHub repositories, data by university of Granada and Internet Movies Firearms Database (IMFDB) imfdb.org. Two approaches are used i.e. sliding window/classification and region proposal/object detection. Some of the algorithms used are VGG16, Inception-V3, Inception-ResnetV2, SSDMobileNetV1, Faster-RCNN Inception-ResnetV2 (FRIRv2), YOLOv3, and YOLOv4. Precision and recall count the most rather than accuracy when object detection is performed so these entire algorithms were tested in terms of them. Yolov4 stands out best amongst all other algorithms and gave a F1-score of 91% along with a mean average precision of 91.73% higher than previously achieved.
29 October 2020 in IEEE Access
Leather defect analysis is important for leather quality grading which directly effects the leather exports. Automated leather sample classification is vital due to slow and subjective nature of the manual process. The major challenges that exist in visual inspection of leather samples for categorization are: the morphology of defects significantly differs and their close examples are not available for transfer learning, unavailability of publicly available data and a benchmark. In this paper, we discuss three important aspects in the identification of industrial leather defects, i.e. the creation of an annotated wet-blue leather image dataset, the transfer of information from different domains to the leather image domain and the design of ensemble networks tailored to the task. We are therefore introducing a new database of wet-blue leather images (Wet-blue Leather Image Dataset (WBLID)) for the classification of defects along with expert annotation data. We proposed a new network EfficientNet-B3+ ResNext-101. The proposed EfficientNet-B3+ ResNext-101 ensemble architecture significantly outperforms all other state-of-the-art methods in terms of AUC and F1-score.