Computer_Vision_Segmentation_Fun

Segmentation of Images and Video.

Tools:

• pytorch

Models:

• Classic model - GrabCut
• Deep model - Deeplabv3_resnet101

Flow:

part 1 - classic model & deep model on images for:

• segmentation
• augmentation
• adverserial example (background) - checked by classification on the object in the image

part 2 - deep model in video:

• using green background from:
• my own video
• combining them to fun result: video or on YouTube

detailed report - report

*This is was created in the scope of Computer Vision 2020 course at the Technion