Conversation
alalek
left a comment
There was a problem hiding this comment.
Great contribution!
Please take a look on notes below.
|
|
||
| ### Portrait segmentation visualization | ||
|
|
||
| <img src="./data/human_image.jpg" width="50%" height="50%"><img src="./data/result_test_human.png" width="50%" height="50%"> |
There was a problem hiding this comment.
Used images should be properly licensed.
Please provide information where did you get this image? (dataset, its license)
| * 2. Preprocess image file and do inference. | ||
| * 3. Postprocess image file and visualization. | ||
|
|
||
| The result is in the path `data/result_test_human.png` . |
There was a problem hiding this comment.
.png files should be avoided in documentation (it is lossless format which provides huge files).
Consider using .jpg image format in documentation.
There was a problem hiding this comment.
What is about the source of used image?
There was a problem hiding this comment.
The source of used image comes from Supervise.ly dataset,it is noted in the last line of the readme~
There was a problem hiding this comment.
Used images should have proper licensing terms. Do you have a link on them?
There was a problem hiding this comment.
The sample image comes from supervisely-ecosystem/persons,it's publicly available and free for academic purposes, and there's no license announcement for the dataset. Do you have any suggestions~
| import cv2 as cv | ||
| import numpy as np |
There was a problem hiding this comment.
Please put numpy import before cv2 (there are several reports where importing cv2 is broken without numpy).
|
@wjj19950828 Friendly reminder. |
|
@alalek,I have updated my PR based on your suggestions. Do you have any suggestions for related test cases? Looking forward to it your reply~ |
| @@ -1,6 +1,6 @@ | |||
| # Run PaddlePaddle model by OpenCV | |||
| # Run PaddlePaddle model using OpenCV | |||
There was a problem hiding this comment.
/cc @jiangjiajun As previous contributor please take a quick look on updates of PaddlePaddle samples.
There was a problem hiding this comment.
I have reviewed this change before, @wjj19950828 fix some grammer problems for my pull request
|
@wjj19950828, |
This is a good suggestion, I will replace it with this picture～ |
|
@alalek Hi~is there anything else that needs to be modified in this PR, can you help me review it again? |
alalek
left a comment
There was a problem hiding this comment.
Well done! Thank you for contribution 👍
…demo Add paddle humanseg demo * fixed onnx resize op bug * add humanseg demo for PaddlePaddle sample * update README.md and flake8 format * update func name * update README.md for enviroment setup * update README.md in the way install paddle2onnx * update README.md * update README.md * add paddleseg in requirements.txt * deal with comments * replace picture
Pull Request Readiness Checklist
See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request
Patch to opencv_extra has the same branch name.
This PR do follow contributes
1、fixed opencv.dnn resize && upsample op,support interp_mode == "half_pixel" situation.if not fixed,the onnx and dnn result has diff.
2、add PaddleSeg Portrait Segmentation demo by OpenCV.
3、fixed paddle_resnet50.py bug,change function name:mobilenetv2 to resnet50.
4、update readme.