in this project,i made some addition in the mask-rcnn which are metion below
object detection for image & video
semantic segmentation for both image and video
youtube download video
image download from url
Working of Model
For Videos:
In this model you just give path of video or any youtube video link it will automatically download video from youtube and put it into model, after that model will give you pretty nice video which include (object detection and mask of each frame)
For Image:
Same case for image you only have to put any picture's url and modle will give you output as it give in each frame of video
About
Mask_RCNN for object detection and semantic segmentation on open cv and pytorch by using mm-detection framework