GithubHelp home page GithubHelp logo

cgan_jittor_landscape's Introduction

Jittor 草图生成风景比赛 SPADE-Jittor

本项目以 CVPR2019 的 SPADE 模型作为基础框架,采用 Jittor 实现,相关论文 为 Semantic Image Synthesis with Spatially-Adaptive Normalization,部分结果展示如下:

image.png

1031937_222ee03f9f_b

简介

本项目为第二届计图挑战赛计图 - 草图生成风景比赛 A 榜的参赛项目,包含了全部的代码实现。本项目的特点是:使用 Jittor 框架以 SPADE 模型为基础进行复线,通过强化学习的方法对原始语义分割图进行处理,生成风景图片。

安装

本项目可在单张 GPU 上运行。

运行环境

  • python >= 3.7
  • jittor >= 1.3.4

安装依赖

执行以下命令安装 python 依赖:

pip install -r requirements.txt

数据准备

请从 比赛官网这里 下载数据训练集以及测试集。请将解压后的数据放入 <root>/data 下。

文件结构组织如下:

SPADE-Jittor
--data
 |--train
   |--imgs
   |--labels
 |--val
   |--labels
--options
--scripts
--spade
--util
__init__.py
datasets.py
pix2pix_model.py
pix2pix_trainer.py
test.py
train.py
README.md
requirements.txt

训练与测试

项目内包含了用于训练模型以及在官方测试集上测试模型的脚本。

训练模型,请执行 scripts 文件夹下的 train.sh

cd scripts
bash train.sh

测试模型,请执行 scripts 文件夹下的 test.sh

cd scripts
bash test.sh

其中的关键参数如下:

--name [the name of current training / test on which trained model] \
--n_epoch [total number of epoch in training] \
--batch_size [batch size of training / testing] \
--lr [learning rate] \
--save_epoch_freq [save the model every $~ epochs] \

其他超参数 --no_instance--preprocess_mode--aspect_ratio 以及 --label_nc,用于适配当前数据集,因而无需更改。 更多超参数及其默认值可参考 options

参考文献

[1] Park T, Liu M Y, Wang T C, et al. Semantic image synthesis with spatially-adaptive normalization[C]//Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2019: 2337-2346.

[2] Zhou X, Zhang B, Zhang T, et al. Cocosnet v2: Full-resolution correspondence learning for image translation[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2021: 11465-11475.

[3] Qu Y, Chen Y, Huang J, et al. Enhanced pix2pix dehazing network[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2019: 8160-8168.

cgan_jittor_landscape's People

Watchers

He Bingxiang avatar

Recommend Projects

  • React photo React

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

  • Vue.js photo Vue.js

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

  • Typescript photo Typescript

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

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo 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.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

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

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.