Skip to content

Latest commit

 

History

History
106 lines (89 loc) · 2.99 KB

README.md

File metadata and controls

106 lines (89 loc) · 2.99 KB
Raw

Team 4

Final Presentation
Final Report

You can access the gifs we created from this link.

Members
  • Mert Yazan
  • Berkay Arpacı
  • Marcos Conde
Members Github
Mert Yazan myazann
Berkay Arpacı BBAKY
Marcos Conde mv-lab

Each week's work is put into its own folder. Check the folders for more detailed information about the code.

Week 1

Main Task:
Implement and use mAP for the object detection of a video, find error in optical flow, visualize the flow.

Tasks:
  1. Bounding box detection by adding noise to ground truth
  2. Calculation of mAP for YOLO3, MaskRCNN, SSD
  3. Optical flow visualization
  4. MSEN and PPEN calculation

Week 2

Main Task:
Find the bounding boxes by determining the objects in the foreground.

Tasks:
  1. Gaussian model for Background Segmentation
  2. Adaptive Gaussian
  3. Trying SOTA Background segmentation approaches
  4. Effect of different color spaces in Gaussian

Week 3

Main Task:
Object Detection and Object Tracking

Tasks:
  1. Try pretrained networks for Object Detection
  2. Finetune a network on our dataset
  3. Object Tracking with Maximum Overlap
  4. Object Tracking with Kalman Filter

Week 4

Main Task:
Optical Flow and MTSC

Tasks:
  1. Optical Flow with Block Matcing
  2. Off-the-shelf Optical Flow
  3. MTSC (Multi Track Single Camera)

Week 5

Main Task:
Multi Track Single Camera (MTSC) Tracking and Multi Track Multi Camera (MTMC) Tracking

Tasks:
  1. Improve the MTSC tracking approach done in the previous week
  2. Implement MTMC tracking

References