Report title: Video Understanding and Analysis

Reporter: Zhang Peixuan Senior Engineer Changchun Boli Electronic Technology Co., Ltd.

Reporting time: 10:10-10:45 am, September 17, 2020

Report location: Tencent Conference ID: 206 372 412

Conference password: 0917

School contact: Jia Jiwei jiajiwei@jlu.edu.cn

Report summary:

In recent years, due to the progress of deep learning technology, many fields in computer vision have been greatly developed, including the field of video analysis. Although many problems in the field such as motion recognition and detection, motion analysis and tracking have not been truly resolved, they are gradually showing huge development space and application prospects. This report will combine the top conferences (such as CVPR, ICCV, ECCV) and technical competitions (such as COCO, Activity-Net, Youtube8M) in the computer vision field in recent years, as well as the open source projects of Facebook and Google, to try to summarize the video analysis field Frontier development.

The report will be divided into three parts:

1. Overview of video understanding and analysis

2. Video understanding and analysis based on single frame image

3. Video understanding and analysis based on sequence images

Brief introduction of the speaker:

Zhang Peixuan, a senior engineer, graduated from Jilin University and is currently the R&D manager of Changchun Boli Electronic Technology Co., Ltd. He has won awards such as senior developer of Sina cloud computing and the most outstanding programmer in the software and information service industry of Jilin Province. He has more than 10 years of R&D and engineering experience in the fields of heterogeneous parallel computing, image and video encoding and decoding, and image and video recognition. He is a code contributor to multiple open source communities such as GIMP, LibJPEG, and x265.