Light

shahabty / pspnet-pytorch Goto Github PK

View Code? Open in Web Editor NEW

33.0 4.0 6.0 129 KB

Implemetation of Pyramid Scene Parsing Network in Pytorch

Python 100.00%

python pytorch pspnet caffe semantic-segmentation

pspnet-pytorch's Introduction

PSPNet-Pytorch

This Repo contains an implemetation of "Pyramid Scene Parsing Network" in Pytorch. Pretrained weights are converted from Official Caffe Repo. The performance is: [acc 0.95196], [acc_cls 0.83369], [mean_iu 0.75547], [fwavacc 0.91172] on Validation set of CityScapes Dataset.

Installation

Install Anaconda3 from Here
Create a Conda Environment:

username@PC:~$ conda env create -f environment.yml

Download Cityscapes Dataset from Here
Download Caffe Pretrained from Here and put it in Caffe-PSPNet folder
Run the code:

username@PC:~$ python main.py

Qualitative

Note

Weight conversion from Caffe to Pytorch is modified from pytorch-semseg.
Preprocessing and loss function is modified from pytorch-semantic-segmentation.
Since the differences are significant, I decided not to add a branch to any of the the above project.

pspnet-pytorch's People

Contributors

Stargazers

Watchers

Forkers

narendoraiswamy ryany1994 atlasml zhangan15 ukdsvl cooper-ji

pspnet-pytorch's Issues

Use custom dataset

@shahabty thank you for the code, I works fine :) I could get good results with the cityscapes dataset. However If I have to use a custom dataset, how can I do that?

[acc 0.95196], [acc_cls 0.83369], [mean_iu 0.75547], [fwavacc 0.91172]

May I know how [acc 0.95196], [acc_cls 0.83369], [mean_iu 0.75547], [fwavacc 0.91172] corresponds to the IoU class, iIoU class, IoU category and iIoU category in the official cityscapes benckmarking?

Thanks in advance!

Cannot reproduce your results

Hi,
I downloaded the pretrained caffemodel and put it in the correct directory, I have no errors at all, but the main script gives me IoU of 8%. Do you have any ideas why? I'm using pytorch 0.4.1.

val set result

Hi,
did you try the pspnet from repo pytorch-semseg (https://github.com/meetshah1995/pytorch-semseg)? I think the model structure in your code is almost the same as theirs, but their model can only produce mIoU of about 52% on val set, and I cannot find any difference in data processing, do you know why?

About mean_std value

Hi,
If I want to finetune this PSPNet, How should I set the transform, especially the mean std value?
Thanks

Recommend Projects

React

A declarative, efficient, and flexible JavaScript library for building user interfaces.
Vue.js

🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
Typescript

TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
TensorFlow

An Open Source Machine Learning Framework for Everyone
Django

The Web framework for perfectionists with deadlines.
Laravel

A PHP framework for web artisans
D3

Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

javascript

JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
web

Some thing interesting about web. New door for the world.
server

A server is a program made to process requests and deliver data to clients.
Machine learning

Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Visualization

Some thing interesting about visualization, use data art
Game

Some thing interesting about game, make everyone happy.

Recommend Org

Facebook

We are working to build community through open source technology. NB: members must have two-factor auth.
Microsoft

Open source projects and samples from Microsoft.
Google

Google ❤️ Open Source for everyone.
Alibaba

Alibaba Open Source for everyone
D3

Data-Driven Documents codes.
Tencent

China tencent open source team.