README.md

Machine Learning Paper Review

Introduction

As a novice researcher in the field of artificial intelligence, I find it important not only to read papers quickly and understand them well to gain insight, but also to implement them in code and review their performance.

In this repository, I will share my code implementations and reviews of various papers in the field of machine learning. My hope is that this repository will be a useful resource for other researchers who want to learn more about the latest developments in the field and implement them in code.

My primary focus is on papers related to computer vision and recommendation systems, but I may expand to other areas in the future. If you have any suggestions or requests for papers to review, please feel free to let me know.

Papers Review

I regret you to inform that this GitHub repository is primarily operated for my own AI study, so the some comments in code and paper review posts on my tech blog are written in Korean.

The code below is a simplified version and typically based on .ipynb files.

---

Fundamental

Convolution

Name	Paper	Paper Review	Code	Code Reference
AlexNet	ImageNet Classification with Deep Convolutional Neural Networks		File
ResNet	Deep Residual Learning for Image Recognition		File
VGG	Very Deep Convolutional Networks for Large-Scale Image Recognition		File

Generative Adversarial Network

Name	Paper	Paper Review	Code	Code Reference
CGAN	Conditional Generative Adversarial Nets		File

---

Computer Vision

Encoding

Name	Paper	Paper Review	Code	Code Reference
Fourier Feature	Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains	Link
InstantNGP	Instant Neural Graphics Primitives with a Multiresolution Hash Encoding		File	Link

Generative Adversarial Network

Name	Paper	Paper Review	Code	Code Reference
StyleGAN	A Style-Based Generator Architecture for Generative Adversarial Networks	Link	File	Link
HoloGAN	HoloGAN: Unsupervised learning of 3D representations from natural images	Link

Implicit Function

Name	Paper	Paper Review	Code	Code Reference
PIFu	Pixel-Aligned Implicit Function for High-Resolution Clothed Human Digitization	Link
PIFuHD	PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution 3D Human Digitization		File	Link

Radiance Field

Name	Paper	Paper Review	Code	Code Reference
NeRF	NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis	Link	File	Link

Renderer

Name	Paper	Paper Review	Code	Code Reference
RenderNet	RenderNet: A deep convolutional network for differentiable rendering from 3D shapes	Link

Transformer

Name	Paper	Paper Review	Code	Code Reference
Vision Transformer	An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale		File

Diffusion

Name	Paper	Paper Review	Code	Code Reference
DDPM	Denoising Diffusion Probabilistic Model	Link	File
DiffEdit	DiffEdit: Diffusion-Based Semantic Image Editing with Mask Guidance		File	Link
NCSN	Generative Modeling by Estimating Gradients of the Data Distribution		File
SDE	Score-Based Generative Modeling through Stochastic Differential Equations		File	Link

---

Recommendation System

Autoencoder

Name	Paper	Paper Review	Code	Code Reference
EASE	Embarrassingly Shallow Autoencoders for Sparse Data	Link	File	Link
Multi-DAE	Variational Autoencoders for Collaborative Filtering		File

Variational AutoEncoder

Name	Paper	Paper Review	Code	Code Reference
Multi-VAE	Variational Autoencoders for Collaborative Filtering		File	Link
RecVAE	RecVAE: a New Variational Autoencoder for Top-N Recommendations with Implicit Feedback		File

Acknowledgments

The code and the review in this repository is based on the original implementation by the authors of each paper. We thank them for releasing their code.

License

This project is licensed under the MIT License - see the LICENSE file for details. Some kinds of codes for paper's implementation are protected under the licenses specified by the authors. In such cases, the source of the code is left together. (Please refer to columns named "Code Reference".)