Yolact++ pytoch environment

Anaconda3 Installation can refer to Deeplabv3+ Environment configuration -Anaconda3 + Pytorch1.8 + Cuda10.1 + opencv3.2.0

Environment configuration

1. First of all to pytorch Create a anaconda A virtual environment , The name of the environment can be determined by yourself , Use here yolact++ As an environment name ：

$ conda create -n yolact++ python==3.8

2. Activate after successful installation yolact++ Environmental Science

$ source activate
$ conda activate yolact++

3. Install... In the created environment pytorch, View this machine cuda edition

$ conda install pytorch torchvision torchaudio cudatoolkit=10.2 -c pytorch
$ conda install cudnn=7.6.5

Be careful ：10.2 Place should be cuda Installation version number of
edit ~./bashrc file , Set to use yolact++ In the environment python3.8

alias python='~/anaconda3/envs/yolact++/bin/python3.8'

4. Update environment variables

$ source ~/.bashrc

5. Activate again yolact++

conda activate yolact++

6. test

python
import torch
torch.cuda.is_available()
torch.cuda.device_count()
torch.cuda.current_device()

Yolact++ Project clone installation

1. Clone and install Yolact++

$ git clone https://github.com/dbolya/yolact.git

2. Install the required libraries
   stay yolact++ Environment , route yolact perform

$ pip install cython
$ pip install opencv-python pillow pycocotools matplotlib

3. install DCN

$ git clone https://github.com/jinfagang/DCNv2_latest.git

4. stay yolact++ Environment , route DCNv2_latest perform

$ cd DCNv2_latest
$ python setup.py build develop
$ python testcuda.py # run examples and gradient check on gpu

The test image

Download test pictures to testfile
Next , Weight file to weights Next

Use yolact_plus_base Model

python eval.py --trained_model=weights/yolact_plus_base_54_800000.pth --score_threshold=0.15 --top_k=15 --image=testfiles/elephant.jpg:testfiles/elephant_seg.jpg

Use yolact_plus_resnet50 Model

python eval.py --trained_model=weights/yolact_plus_resnet50_54_800000.pth --score_threshold=0.15 --top_k=15 --image=testfiles/elephant.jpg:testfiles/elephant_seg2.jpg

Test video

python eval.py --trained_model=weights/yolact_plus_resnet50_54_800000.pth --score_threshold=0.15 --top_k=15 --video_multiframe=4 --video=testfiles/drive.mp4:testfiles/drive1.mp4