Deep Learning

Mask R-CNN balloon Training

전두선 2020. 2. 21. 22:54

Mask RCNN

gitbub: https://github.com/jeonds1127/Mask_RCNN


> Instance Segmentation

  • Classification: There is a balloon in this image.
  • Semantic Segmentation: These are all the balloon pixels.
  • Object Detection: There are 7 balloons in this image at these locations. We’re starting to account for objects that overlap.
  • Instance Segmentation: There are 7 balloons at these locations, and these are the pixels that belong to each one.

> balloon.py

balloon dataset: https://github.com/matterport/Mask_RCNN/releases

pretrain: https://drive.google.com/open?id=1K7RIMHNeVCnWpIE04s8qPh33m367MHLZ

 

Training

$ python balloon.py train --dataset='<add>/balloon' --weights=coco --logs='<add>/weights'

Testing

$ python balloon.py splash --weights='<add>/mask_rcnn_balloon_0030.h5' --image='<add>/balloon/val/410488422_5f8991f26e_b.jpg'

 

 

'Deep Learning' 카테고리의 다른 글

CNN을 활용한 Traffic Sign Detection  (0) 2020.02.26
Detection algorithm  (0) 2020.02.06
Video Stabilization Algorithm  (0) 2020.02.06