視訊串流與追蹤

This course aims at introducing video streaming over Internet and its applications such as surveillance, remote object tracking etc. The technologies involved include video coding, streaming protocol, object detection and localization, object classification and identification, object tracking, etc. Students will learn both machine learning methods as well as conventional ways on these technologies.

教師未提供此項資料

無備註

教師未提供此項資料

Mid-terma + final exam: 40% Programming homeworks (4~5) : 45% Paper presentation: 15%

• Ch1. Overview on video streaming and its applications

  講授：

  3

• Ch2. Introduction to Deep Neural network and its variations

  - CNN, RNN

  講授：

  3

• Ch3. Image Classification

  - AlexNet, ZFnet, VGG, GoogLeNet, ResNet, and some variations

  講授：

  3

• Ch4. Object Detection and Localization

  - Overfeat, R-CNN, SPPnet, Fast R-CNN, Faster R-CNN, SDD, YOLO, RetinaNet, FPN

  講授：

  3

• Ch5. Object Segmentation

  - FCN, DeepLab, SegNet, E-Net, ICnet,DeepMask, SharpMask, I-FCN, R-FCN, MNC, Mask R-CNN

  講授：

  3

• Ch6. Object Representation and Identification

  - DeepFace, FaceNet, BinaryHashingCode, DeepBit

  講授：

  3

• Ch7. Object Tracking (I) - video object tracking (VOT)

  - RNN-based tracking: RATM, ROLO -CNN-based tracking: GOTURN, HCF, RFH, SiamFC, SiamRPN, SiamRPN++, SiamMask

  講授：

  6

• Ch8. Object tracking (II) - video object segmentation (VOS)

  PML, RGMP, FEELVOS, RANET, TVOS, STM

  講授：

  3

• Ch9. Video coding

  - spatial coding - temporal coding - H.264 - HEVC - Audio coding

  講授：

  9

• Ch10. Streaming protocol

  -ES, PES, PS, TS - DVB IPTV, HTTP Live Streaming (HLS), Adobe Media Server

  講授：

  6

hand-on lecture notes

地點

教師未提供此項資料

時間

教師未提供此項資料

聯絡方式

教師未提供此項資料