Proceedings of the 4th International Symposium on Signal Processing Systems最新文献_第2页

A Lightweight Video Summarization Method Considering the Subjective Transition Degree for Online Educational Screen Content Videos 一种考虑主观过渡程度的在线教育屏幕内容视频轻量级摘要方法

Proceedings of the 4th International Symposium on Signal Processing Systems Pub Date : 2022-03-25 DOI: 10.1145/3532342.3532354

Qinqin Meng, Kaifang Yang, Yanchao Gong

{"title":"A Lightweight Video Summarization Method Considering the Subjective Transition Degree for Online Educational Screen Content Videos","authors":"Qinqin Meng, Kaifang Yang, Yanchao Gong","doi":"10.1145/3532342.3532354","DOIUrl":"https://doi.org/10.1145/3532342.3532354","url":null,"abstract":"With the popularization of online education, the number of educational videos is increasing, and some shoddy and dangerous videos also become potential threat to the mental health of students and the public security. Therefore, efficient video summarization technology is critical for the analysis, retrieval, and management of online educational videos. Online education videos usually belong to the screen content videos (SCV), and is typical real-time communication system, which urgently needs lightweight technologies with fast speed and low hardware requirements. Therefore, the traditional video summarization method for nature videos or with high complexity cannot be effectively applied for the online educational videos. SCV generated by screen recording the play of the Power Point (PPT-SCV) has also been widely applied in education field. Therefore, taking the PPT-SCV as an example, this paper proposed a lightweight video summarization method for the educational videos. First, the average standard deviation among frames and the variance of frame difference in a video clip which can effectively reflect the content characteristics of PPT-SCV were used to obtain the category of transitions and the key frames. Then, the subjective transition degree in video is used for marking the knowledge importance of each key frames, and a lightweight video summarization method was finally proposed. Experimental results demonstrated that the proposed method can efficiently locate the key frames which is in line with subjective perception.","PeriodicalId":398859,"journal":{"name":"Proceedings of the 4th International Symposium on Signal Processing Systems","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121471255","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An Irregularly Dropped Garbage Detection Method Based on Improved YOLOv5s 基于改进YOLOv5s的不规则丢弃垃圾检测方法

Proceedings of the 4th International Symposium on Signal Processing Systems Pub Date : 2022-03-25 DOI: 10.1145/3532342.3532344

Yi Zhan, Yuanping Xu, Chaolong Zhang, Zhijie Xu, Benjun Guo

{"title":"An Irregularly Dropped Garbage Detection Method Based on Improved YOLOv5s","authors":"Yi Zhan, Yuanping Xu, Chaolong Zhang, Zhijie Xu, Benjun Guo","doi":"10.1145/3532342.3532344","DOIUrl":"https://doi.org/10.1145/3532342.3532344","url":null,"abstract":"Waste sorting and recycling play a significant role in carbon neutrality, and the government has promoted waste sorting stations in various cities while the stations have limited efficiency due to the absence of intelligent surveillance systems to monitor and analyze the scene in waste stations, especially to detect the irregularly dropped garbage. To take the most advantage of these stations, this study proposes an improved YOLO (You Only Look Once) v5s detector named YOLOv5s-Garbage to monitor waste sorting stations in real-time. This study enhances its ability to detect garbage by introducing CBAM (Convolutional Block Attention Module) and using EIoU (Efficient Intersection over Union) to accelerate the convergence of the bonding box loss. According to experiments, the mAP of YOLOv5s-Garbage on the waste sorting dataset reaches 89.7%, which is 3.3% higher than the classical YOLOv5s. This study then combines the DeepSort tracking algorithm and re-filter process to filter the target garbage to distinguish the irregularly dropped garbage and normal one, which reduces the false alarm significantly.","PeriodicalId":398859,"journal":{"name":"Proceedings of the 4th International Symposium on Signal Processing Systems","volume":"100 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124116439","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 3

Uni-Dimensional Autoencoder Reinforced Multilayer Perceptron Network for Individual Behavior Detection 用于个体行为检测的一维自编码器增强多层感知器网络

Proceedings of the 4th International Symposium on Signal Processing Systems Pub Date : 2022-03-25 DOI: 10.1145/3532342.3532353

Lingzhe Wang, Yuefan Hao, Ying Liu

引用次数: 0

Bilateral Pose Transformer for Human Pose Estimation 用于人体姿态估计的双边姿态转换器

Proceedings of the 4th International Symposium on Signal Processing Systems Pub Date : 2022-03-25 DOI: 10.1145/3532342.3532346

Chia-Chen Yen, Tao Pin, Hongmin Xu

{"title":"Bilateral Pose Transformer for Human Pose Estimation","authors":"Chia-Chen Yen, Tao Pin, Hongmin Xu","doi":"10.1145/3532342.3532346","DOIUrl":"https://doi.org/10.1145/3532342.3532346","url":null,"abstract":"Human Pose is a well-defined fundamental task researched by the computer vision community for years. Previous Convolutional Neural Network (CNN) based works have achieved significant success in the human pose. Recently, Vision Transformer (VT) has shown superior performance on computer vision tasks. However, current VT methods emphasize local information less and often focus on only a single scale feature that may not be suitable for dense image prediction tasks, which essentially requires multi-scale representations. In this paper, we propose a novel Bilateral Pose Transformer (BPT) framework to handle the human pose. Specifically, BPT consists of an innovated bilateral branch encoder and a multi-scale integrating decoder. The bilateral branch encoder contains a Context Branch (CB) and Spatial Branch (SB). The CB involves a VT-based backbone to capture the context clues and produce multi-scale context features. The CNN-based SB maintains high-resolution representations containing rich spatial information to introduce the local spatial information that supplements the CB explicitly. About the decoder, a Mixed Feature Module consisting of local attention CNN is proposed to integrate the various-scale context and spatial features effectively. Experiments demonstrate that our approach achieves competitive performances in human pose estimation. Specifically, compared to the HRNet [1], the BPT saves 43% GFLOPs and drops only 0.1 points AP, achieving 75.7% AP with 9.0 GFLOPs, on the COCO keypoints dataset.","PeriodicalId":398859,"journal":{"name":"Proceedings of the 4th International Symposium on Signal Processing Systems","volume":"24 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120993604","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Robust and Efficient Shoe Print Image Retrieval Using Spatial Transformer Network and Deep Hashing 基于空间变压器网络和深度哈希的鲁棒高效鞋印图像检索

Proceedings of the 4th International Symposium on Signal Processing Systems Pub Date : 2022-03-25 DOI: 10.1145/3532342.3532356

Wei Liu, Dawei Xu

{"title":"Robust and Efficient Shoe Print Image Retrieval Using Spatial Transformer Network and Deep Hashing","authors":"Wei Liu, Dawei Xu","doi":"10.1145/3532342.3532356","DOIUrl":"https://doi.org/10.1145/3532342.3532356","url":null,"abstract":"In recent years, great progress has been made on the topic of shoe print image retrieval. However, it still remains a big challenge to accurately retrieve well-matched shoe print images from a huge database in real time. Deep hashing method has been proved to be effective in large-scale image retrieval in many applications. Output of deep hash network can be represented as a binary bit hash code, which helps to reduce storage space and retrieval time. Existing shoe print image retrieval methods have poor performance in instance retrieval because of shortage of classification information, sufficient sample information and feature quantization information. In order to overcome the problem, we put forward an end-to-end network to learn short deep hash codes. The learned hash code preserves the classification and small sample information very well. Moreover, due to the fact that STN (Spatial Transformer Network) block is simultaneously embedded into the hash network to enhance the retrieval ability for rotated shoe print images, the problem of rotation misalignment can be solved and the retrieval accuracy is improved. Furthermore, in order to make better use of class label information, we presented a new joint loss function. This loss function helps the network map both images’ classification information and similarities into hash codes and reduce quantitative loss. In addition, we used triple labels to alleviate sample imbalance problem. Experiments on database including 10,500 shoe print images show that our proposed method can improve the retrieval performance. The proposed approach can yield a mAP (mean Average Precision) of 0.83 and a recall of 0.35, which demonstrates the discriminatory power of the learned hash codes in shoe print image retrieval application.","PeriodicalId":398859,"journal":{"name":"Proceedings of the 4th International Symposium on Signal Processing Systems","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125443763","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 1

Suppress Target Height Induced False Alarm in CSAR Moving Target Detection 抑制CSAR运动目标检测中目标高度引起的虚警

Proceedings of the 4th International Symposium on Signal Processing Systems Pub Date : 2022-03-25 DOI: 10.1145/3532342.3532350

Fan Ding, Wenjie Shen, Y. Li, Yun Lin, Yanping Wang

引用次数: 0

Proceedings of the 4th International Symposium on Signal Processing Systems 第四届信号处理系统国际研讨会论文集

Proceedings of the 4th International Symposium on Signal Processing Systems Pub Date : 1900-01-01 DOI: 10.1145/3532342

引用次数: 0