wonbeomjang / esrgan-pytorch

re-implementation super resolution pytorch using ESRGAN

Hi, There 👋

Wonbeom Jang

💻 Work Experience

Role	Company	Period
ML/DL Engineer	SK Telecom	2024.01. ~
Deep Learning Researcher (Intern)	Neurocle	2023.01. ~ 2023.08.

🔬 Research Experience

Role	Company	Period
Undergraduate Researcher	Computer Vision Machine Learning Lab. (Advisor. Prof. Junseok Kwon)	2019.08. ~ 2020.08.

📄 Paper

• Deep-plane sweep generative adversarial network for consistent multi-view depth estimation, Machine Vision and Applications, 2022, Dong Wook Shu, Wonbeom Jang, Heebin Yoo, and Hong-Chang Shin, Junseok Kwon

🛠️ Tech Skill

Classification	Skills
Machine Learning
Language - Favorite
Language - Useable
ETC

🏅 Awards

Awards	Grade	Date	Note
Military Service Open SW Online Hackathon	Navy Chief of Staff Award	2021.10.	Canary
Barrier-free App Development Contest	Encouragement Award	2020.11.	Zipyak

🎓 Education

• Bachelor of Engineering in Computer Science & Engineering, Chung-Ang University 2018. 03. ~ 2024.02.
  • GPA : 4.41/4.5

🚂 Toy Project

2021

• Canary: Image mosaic for military AI_APP_WEB_Canary_Canary_2021, yolov5-knowledge-distillation
  • Proceed with knowledge distillation in yolov5
  • Analyze problems of model and data, and improve performance over 14.9%p
  • Building an administrator server for log collection and deployment of the latest models
  • Some part of codes receives 23 stars at github

2019

• Zipyak: Pharmacological Information Service for the Blind link
  • Use google cloud vision OCR to iOS application
  • Creating Pharmacological Information Retrieval Algorithm with OCR Data
  • User interviews and user-friendly app design
• Photo Filter Application – ESRGAN-pytorch, ESRGAN-pytorch, ESRGAN-pytorch
  • Implementing Style transfer using cnn, CycleGAN, ESRGAN
  • Deploy using pip and github to use on Django
  • ESRGAN receive 75 stars at github
• Fash: Fashion classification and recommendation system
  • Build classification model and deploy to iOS applications using libtorch and object c++

2018

• Hair segmentation for mobile device - github.com/wonbeomjang/mobile-hair-segmentation-pytorch
  • Implementing HairMatteNet and Improving Performance
  • Improve performance and reduce execution time by changing model structure and quantization
  • Improve IOU 0.73%p, reduce interference time 42.9% and model size 56%
  • This project receives 102 stars at github

📧 Contact me

[email protected]

esrgan-pytorch's Issues

损失函数相关

我看到您loss.py中perception_loss与SRGAN中相同，但是ESRGAN中作者指出该损失应该取激活之前的特征进行求得，请问根据您使用的loss情况，超分效果如何？指标是否下降？

Output images are completely black

Dear author,
Thank you so much for sharing the simplified code of ESRGAN.

I have used your training code for my own dataset, but testing code produced completely black super resolved images.

Kindly guide.
Best,
Farooq

code exit

when running 'python main.py', this program will be exit right now . and no message output

How to use the src/test.py?

in src, python3 test.py in my PC, no results. and how to use it ?

wanna the weight file

Can you share the trained weight file? My result is not very good

Dimension mismatch error on changing image dimensions

My hr images are 256x256 and lr is 64x64, with batch_size=8 and scale=4

I'm getting this error:
RuntimeError: size mismatch, m1: [8 x 73728], m2: [8192 x 100] at /opt/conda/conda-bld/pytorch_1579022060824/work/aten/src/THC/generic/THCTensorMathBlas.cu:290

Where is the Discriminator?

How much GPU memory is needed for training?

Hi，
I set the batchsize to 1 to train on a RTX 3080, but it prompts that the GPU memory is insufficient to train. So I would like to ask, what GPU are you training on?
Sincerely，
Chison