Abstract: Recently few-shot segmentation (FSS) of metal surface defects has been extensively developed. Most previous methods strive to achieve generalization through the pixel-based prototypes ...

Abstract: Interpreting extended, untrimmed videos by dividing them into meaningful action segments is fundamental for automated analysis of human activities. While modern models can represent ...

The goal of the project is to inference a deep learning semantic segmentation model on a webcam video stream in Linux in order to remove background and provide a clean video stream with a person only.

This project demonstrates instance segmentation using Mask R-CNN with the OpenCV DNN module. The model is pre-trained on the COCO dataset and can detect and segment multiple object classes in images.