JRL‐YOLO: A Novel Jump‐Join Repetitious Learning Structure for Real‐Time Dangerous Object Detection

Abstract

Campus security incidents occur from time to time, which seriously affect the public security. In recent years, the rapid development of artificial intelligence has brought technical support for campus intelligent security. In order to quickly recognize and locate dangerous targets on campus, an improved YOLOv3-Tiny model is proposed for dangerous target detection. Since the biggest advantage of this model is that it can achieve higher precision with very fewer parameters than YOLOv3-Tiny, it is one of the Tinier-YOLO models. In this paper, the dangerous targets include dangerous objects and dangerous actions. The main contributions of this work include the following: firstly, the detection of dangerous objects and dangerous actions is integrated into one model, and the model can achieve higher accuracy with fewer parameters. Secondly, to solve the problem of insufficient YOLOv3-Tiny target detection, a jump-join repetitious learning (JRL) structure is proposed, combined with the spatial pyramid pooling (SPP), which serves as the new backbone network of YOLOv3-Tiny and can accelerate the speed of feature extraction while integrating features of different scales. Finally, the soft-NMS and DIoU-NMS algorithm are combined to effectively reduce the missing detection when two targets are too close. Experimental tests on self-made datasets of dangerous targets show that the average MAP value of the JRL-YOLO algorithm is 85.03%, which increases by 3.22 percent compared with YOLOv3-Tiny. On the VOC2007 dataset, the proposed method has a 9.29 percent increase in detection accuracy compared to that using YOLOv3-Tiny and a 2.38 percent increase compared to that employing YOLOv4-Tiny, respectively. These results all evidence the great improvement in detection accuracy brought by the proposed method. Moreover, when testing the dataset of dangerous targets, the model size of JRL-YOLO is 5.84 M, which is about one-fifth of the size of YOLOv3-Tiny (33.1 M) and one-third of the size of YOLOv4-Tiny (22.4 M), separately.

TensorCrypto: High Throughput Acceleration of Lattice-Based Cryptography Using Tensor Core on GPU

“…Although this paper only use the proposed tensor-corebased polynomial convolution for cryptography, it can be extended to support other applications on GPU. For instance, polynomial convolution is used to perform image reconstruction [17], feature extraction [18], data preprocessing [19] and signal processing [20]. These applications may also benefits from the tensor-core-based polynomial convolution to achieve high throughput performance.…”

Material identification of construction machinery based on multisource sensor information fusion

“…Considering that in complex environments (rain and fog, dark and weak environments, etc. ), as the visual sensors are susceptible to environmental interference, resulting in poor imaging quality and blurred vision, leading to the lack of information on object features, the above algorithms are of great significance in improving the recognition accuracy of objects in complex environments by studying multisource sensor fusion, which provides a reference idea for the research in this paper, but most of the above algorithms combine the surface features of objects However, as most of the above algorithms combine the surface features of the object for detection and recognition, when the camera is affected by dark and weak light, the color, texture, morphology, and other features of the object are not obvious, which will greatly affect the detection accuracy of the algorithm, the paper combines YOLO has strong feature extraction capability, it can effectively obtain the shallow features and deep semantic features of the object 9–12 . Therefore, based on the analysis of the advantages of YOLO, it is introduced into the paper for migration learning of construction machinery materials, and then the integrated multisource information is used to comprehensively infer the category of objects, which will become the key to the autonomous and successful excavation of materials by the loader, and will strongly enhance the intelligence of the loader.…”