2021 International Conference on Computer Communication and Artificial Intelligence (CCAI)最新文献

A Display Method of Large Underwater Photo-Mosaics Based on Pyramid Model of Tiled Images 一种基于图像金字塔模型的大型水下照片拼接显示方法

2021 International Conference on Computer Communication and Artificial Intelligence (CCAI) Pub Date : 2021-05-07 DOI: 10.1109/CCAI50917.2021.9447465

Nannan Liu, Xiaoming Li

{"title":"A Display Method of Large Underwater Photo-Mosaics Based on Pyramid Model of Tiled Images","authors":"Nannan Liu, Xiaoming Li","doi":"10.1109/CCAI50917.2021.9447465","DOIUrl":"https://doi.org/10.1109/CCAI50917.2021.9447465","url":null,"abstract":"In deep seafloor exploration, optical imaging provides short-range, high resolution visual information. In order to obtain a wide range of detailed visual information, it is common to stitch multiple images into a photo-mosaic which could reach tens of billions of pixels in size. Due to hardware and software constraints, it is very difficult even impossible to browse and display such a large image with existing image viewers. In this paper, we propose a display method based on pyramid model and develop a super large image display system of tiled images dedicated to underwater photo-mosaics display. Before the image can be displayed, the first step is to construct and store the multi-resolution hierarchical model of tiles, in which sub-mosaics of the same size are original tiles with the highest resolution. Similar to quadtree coding, each tile is encoded based on its location in the pyramid. Then all tile data are stored in MongoDB database. Each record in MongoDB is a key-value pair structure corresponding to a specific tile, in which key is the encoding of a tile and value is the tiled image data stored in binary stream format. Image display is implemented based on graphical user interface. By mouse operation, the system can display images at different resolutions and browse different part of the image. Based on the pyramid model and key-value storage structure, there are more than 60,000 high-definition tiled images in our application. The full panorama resolution is 16.2 billion pixels, about 45GB in RAM, but only the tiles displayed in the current window need to be loaded. Hence, our system reduces the memory requirement greatly and makes image browsing more smoothly.","PeriodicalId":121785,"journal":{"name":"2021 International Conference on Computer Communication and Artificial Intelligence (CCAI)","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115705019","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

A Correlation Filter Tracking Algorithm Incorporating Time Information 一种结合时间信息的相关滤波跟踪算法

2021 International Conference on Computer Communication and Artificial Intelligence (CCAI) Pub Date : 2021-05-07 DOI: 10.1109/CCAI50917.2021.9447479

Xiaoshuo Jia, Shangyou Zeng

引用次数: 0

Simple-NMS: Improved Pedestrian Detection with New Constraints Simple-NMS:基于新约束的改进行人检测

2021 International Conference on Computer Communication and Artificial Intelligence (CCAI) Pub Date : 2021-05-07 DOI: 10.1109/CCAI50917.2021.9447459

Li Tian, Zhaogong Zhang

{"title":"Simple-NMS: Improved Pedestrian Detection with New Constraints","authors":"Li Tian, Zhaogong Zhang","doi":"10.1109/CCAI50917.2021.9447459","DOIUrl":"https://doi.org/10.1109/CCAI50917.2021.9447459","url":null,"abstract":"Non-maximum suppression (NMS) is an indispensable post-processing step in object detection. After a picture is detected by the target, a large number of redundant anchors will be obtained, so we need to go through NMS to process these repeated anchors. Its essence is to extract the target detection frame with high confidence and suppress the false detection frame with low confidence. The biggest problem in the NMS algorithm is that it completely removes adjacent low-confidence detection frames. In some images with dense targets, this is very likely to cause missed detection and false detection. Therefore, we proposed the Simple-NMS algorithm, which adds two new thresholds to the original NMS, and sets a constraint condition different from the original based on the new threshold, and does not change the complexity of the traditional NMS algorithm, which can be very good Improve the effectiveness of NMS missed targets. The new NMS algorithm is improved on the standard data sets PASAL VOC2007 and MS COCO. In addition, the algorithm is simple and efficient, and can be easily integrated into any other object detection process.","PeriodicalId":121785,"journal":{"name":"2021 International Conference on Computer Communication and Artificial Intelligence (CCAI)","volume":"182 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114139502","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

A Satisfactory Detection Framework on Online-shopping Comments 一个满意的网络购物评论检测框架

2021 International Conference on Computer Communication and Artificial Intelligence (CCAI) Pub Date : 2021-05-07 DOI: 10.1109/CCAI50917.2021.9447490

Hanqing Liu

引用次数: 0

A Collaborative Cyberinfrastructure for Research and Education Using Geospatial Model 基于地理空间模型的研究与教育协同网络基础设施

2021 International Conference on Computer Communication and Artificial Intelligence (CCAI) Pub Date : 2021-05-07 DOI: 10.1109/CCAI50917.2021.9447513

Wei Wan, Xinwei Zhao, Xingqiang Du, Zhenkun Yang

{"title":"A Collaborative Cyberinfrastructure for Research and Education Using Geospatial Model","authors":"Wei Wan, Xinwei Zhao, Xingqiang Du, Zhenkun Yang","doi":"10.1109/CCAI50917.2021.9447513","DOIUrl":"https://doi.org/10.1109/CCAI50917.2021.9447513","url":null,"abstract":"Many geospatial modelers around the world use remote sensing data to simulate environmental processes, ecological system and testing agricultural management scenarios. Once these tasks are complete including publication of results, the models generally are not published or made available to the public for further use and improvement. Sharing of models may open doors for new collaborations, and avoids duplication of efforts. For researchers, who are interested in sharing models, there are limited avenues to publishing their models to the wider community. Towards filling this gap, a prototype cyberinfrastructure is developed for publishing, sharing and running geospatial models in an interactive GIS-enabled web environment. Users can utilize the Hub to publish or upload their own models, search and download existing models or data developed by others, run simulations including processing, analysis and calibration using computing and storage resources provided by Cloud. This should open the way for easy development of a variety of web-enabled tools for probing and presenting digital geospatial data in ways that can help addressing issues around the globe.","PeriodicalId":121785,"journal":{"name":"2021 International Conference on Computer Communication and Artificial Intelligence (CCAI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129688169","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

Exploration on Channel-interactive Features in Silent Speech Recognition 无声语音识别中通道交互特性的探索

2021 International Conference on Computer Communication and Artificial Intelligence (CCAI) Pub Date : 2021-05-07 DOI: 10.1109/CCAI50917.2021.9447446

Ming Zhang, You Wang, Guang Li, Xiang Sui

引用次数: 0

Research on Video Fire Detection Algorithm Based on Attention Mechanism 基于注意机制的视频火灾检测算法研究

2021 International Conference on Computer Communication and Artificial Intelligence (CCAI) Pub Date : 2021-05-07 DOI: 10.1109/CCAI50917.2021.9447460

Wenting Ouyang, Y. Fang, Aimin Xiong, Haofei He

引用次数: 1

Considering Filter Importance and Irreplaceability In Filter Pruning 考虑滤波器重要性和不可替代性的滤波器剪枝

2021 International Conference on Computer Communication and Artificial Intelligence (CCAI) Pub Date : 2021-05-07 DOI: 10.1109/CCAI50917.2021.9447516

Wei Lu

引用次数: 0

Tai Le Character Recognition Using Deep Convolution Neural Networks 基于深度卷积神经网络的汉字识别

2021 International Conference on Computer Communication and Artificial Intelligence (CCAI) Pub Date : 2021-05-07 DOI: 10.1109/CCAI50917.2021.9447467

Jing-ying Zhao, Hai Guo, Yifan Liu, N. Dong, Xinyue Wang

引用次数: 1

Lightweight Real-time Object Detection Model for UAV Platform 无人机平台轻量化实时目标检测模型

2021 International Conference on Computer Communication and Artificial Intelligence (CCAI) Pub Date : 2021-05-07 DOI: 10.1109/CCAI50917.2021.9447518

Xueli Wang, Xingtao Zhuang, Wei Zhang, Yunfang Chen, Yanchao Li

{"title":"Lightweight Real-time Object Detection Model for UAV Platform","authors":"Xueli Wang, Xingtao Zhuang, Wei Zhang, Yunfang Chen, Yanchao Li","doi":"10.1109/CCAI50917.2021.9447518","DOIUrl":"https://doi.org/10.1109/CCAI50917.2021.9447518","url":null,"abstract":"Real-time detecting objects on captured images on UAV (Unmanned Aerial Vehicle) platforms, rather than barely transmitting images back to supporting equipment for post-processing, is a core requirement for advanced UAV applications. However, due to limited computing capacity and memory of UAV platforms, it is very challenging to deploy real-time detection models on them. In addition, there are more small objects in aerial images, which makes it more difficult to detect accurately. To solve these problems, this paper brings dense connection to Yolo(You Only Look Once)v3 network, and proposes Yolo-LiteDense model. The backbone of Yolo-LiteDense is densely connected, which improves the performance of feature extraction. Then, we enforce channel pruning to Yolo-LiteDense model by pruning less informative channels with less scaling factors. After pruning, parameters and weight size of the model are compressed significantly, and inference time is also shortened. Evaluation results on VisDrone2018-DET show that parameters and weight size of Yolo-LiteDense are 83% and inference time is 30% less than Yolov3-SPP with comparable average precision. In addition, this paper also proposes the lighter version of Yolo-LiteDense, Yolo-DenseNano. Parameters and weight size of Yolo-LiteDense are 70% less than Yolov3-tiny with 2.68 times greater average precision.","PeriodicalId":121785,"journal":{"name":"2021 International Conference on Computer Communication and Artificial Intelligence (CCAI)","volume":"85 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127662283","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