Skip to content

guiyan2018/Learning-Interactive-Multi-Object-Segmentation

 
 

Repository files navigation

Learning Interactive Multi-Object Segmentation

1. Paper: "Yan Gui, Bingqing Zhou, Jianming Zhang, Cheng Sun, Lingyun Xiang, Jin Zhang. Learning Interactive Multi-Object Segmentation through Appearance Embedding and Spatial Attention, Submitted to IET Image Processing, 2021"

2. Run Demo App

2.1 Download model file, and put it to models folder.

2.2 config python env,install dependent packages and run deom .

## 1. create conda virtual env.
conda create -n mos python=3.6

## 2. activate conda virtual env.
conda activate mos

## 3. install pytorch, reference url: https://pytorch.org.
conda install pytorch torchvision torchaudio cudatoolkit=10.2 -c pytorch

## 4. install other dependent packages.
conda install numpy matplotlib pillow opencv-python

## 5. into the workspaces of demo.
cd ./demo

## 6. run demo app by python file. (if you use ssh connect linux server to run deme app, you can skip this step, see 7-th step).
python demo.py

## 7. run demo app by jupyter notebook (you need run `conda install -c conda-forge notebook` to install jupyter notebook), and then run the last cell of `Demo.ipynb`.

How to Segmentation, you can see chapter 2.3

2.3 Segmentation Demo

operation:

  • mouse:
    • [left button]:interacte
    • [right button]:cancel last interactation
  • keyboard:
    • [number key, include 1-9]: n-th object mark
    • ['p' key]: predict result when not in "auto predict" mode
    • ['ctrl' + 'alt' + 's' key]:save result inlcude predict label, embedding map(random projection), visual attention map
    • ['c' key]: change mode, 'auto predict' or 'press 'p' to predict'
    • ['b' key]: change to before image
    • ['a' key]: change to after image

example

====================================================

Email: bingqiangzhou@qq.com (Bingqiang Zhou)

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 100.0%