Video-Object-Detection-

Instalation

1. Download and place INSTRUCTIONS.sh in a directory of your choice

   • First option: Code (green top right button) -> Dowload ZIP -> Place and unzip the file in a folder.
   • Second option: git clone https://github.com/ga1an/Video-Object-Detection-.git in a terminal place in the directory.

2. Move to the downloaded folder:

   cd Video-Object-Dectection-/

3. Give right permisions to the bash file INSTRUCTIONS.sh:

   chmod +x INSTRUCTIONS.sh

4. Run INSTRUCTIONS.sh:

   ./INSTRUCTIONS.sh

   • Two confimation steps will appear. Write 'y' on the terminal in order to continue the instalation in both cases.

5. Donwload and copy .pth inside model.pytorch/ directory:

   baseline: https://drive.google.com/file/d/1W17f9GC60rHU47lUeOEfU--Ra-LTw3Tq/view?usp=sharing

   MEGA: https://drive.google.com/file/d/1ZnAdFafF1vW9Lnpw-RPF1AD_csw61lBY/view?usp=sharing

6. Dowload and unzip additional matirials in model.pytorch/datasets directory:

   Can be found in the moodle page of the subject-> LABS -> Lab 2: Video object detection -> LAB2-Session1-Aditional material (https://posgrado.uam.es/mod/resource/view.php?id=908972)

   Second option: https://drive.google.com/file/d/1HWknu9savYKZBne2pfLCPT9LlGRHnLXm/view?usp=sharing

7. activate MEGA enviroment and move to mega.pytorch/ folder:

   conda activate MEGA

   cd mega.pytorch/

8. Now you should be able to run the demo:

   python demo/demo.py base configs/vid_R_101_C4_1x.yaml R_101.pth --suffix ".JPEG" --visualize-path datasets/image_folder/ --output-folder visalization

9. Running base and mega:

• Base:

python demo/demo.py base configs/vid_R_101_C4_1x.yaml R_101.pth --video --visualize-path datasets/video.avi --output-folder visualization [--output-video]

• Mega:

python demo/demo.py mega configs/MEGA/vid_R_101_C4_MEGA_1x.yaml MEGA_R_101.pth --video --visualize-path datasets/video.avi --output-folder visualization [--output-video]

Remeber to check that the .pth paths and the video paths are coherent with the ones on your directory.

Additional information

Aditional information can be found in the original GitHub repository (https://github.com/Scalsol/mega.pytorch/tree/master).

• Other models weights: https://github.com/Scalsol/mega.pytorch/blob/master/README.md - section Main Results

• Requeriments: https://github.com/Scalsol/mega.pytorch/blob/master/INSTALL.md