Computer-Vision-Paper-Reviews

Computer Vision Paper Reviews with Key Summary along Papers & Codes.

Jonathan Choi 2021

The repository provides 100+ Papers across Computer Vision fields converted as Jupyter Notebook, with the Key Summary and End to End Code Practice.

---

Contents

The goal of the repository is providing an end to end study scripts of most read and important papers.

The prefered readers are not limited for researchers, but also for students and engieeners from beginner to the professions in computer vision fields .

To provide the rich understanding, each paper contains following three main contents.

Key Summary: Providing key summaries and terminologies of the paper so that even beginner can study easily and perfectly.

Code Practice: Providing an end to end study script of codes for the paper.

Jupyter Notebook edited Original Paper: Providing the Jupyter notbook converted Original Paper for easy and fast modification and understanding.

---

Category/Paper/

Paper_Review_Practice.ipynb includes

Key Summary according to the flow of Original Paper (Jupyter Notebook Edited) with the End to End Code Practice

Paper.ipynb includes

Original Paper (Jupyter Notebook Edited)

Review.ipynb includes

Key Summary

Practice.ipynb includes

End to End Code Practice

---

Visual Correspondence

NCNET - Neighbourhood consensus networks for estimating image correspondences

• Paper & Code Review

• Original Paper

WeakAlign - End-to-end weakly-supervised semantic alignment

• Paper & Code Review

• Original Paper

---

Transformers

[DETR] End-to-End Object Detection with Transformers

• Original Paper

[Vision Transformer] An Image Is Worth 16x16 Words: Transformers For Image Recognition at Scale

• Original Paper

Swin Transformer: Hierarchical Vision Transformer using Shifted Windows

• Original Paper
• Official Github

[Transformer] Attention Is All You Need

• Original Paper

Vision Transformers for Dense Prediction

• Original Paper

---

Object Detection

Feature Pyramid Networks for Object Detection

• Original Paper

Selective Search for Object Recognition

• Original Paper

R-CNN

• Original Paper

Fast R-CNN

• Original Paper

Faster R-CNN

• Original Paper

Sparse R-CNN

• Original Paper

YOLOv4: Optimal Speed and Accuracy of Object Detection

• Paper & Code Review
• Original Paper

---

Segmentation

Panoptic Feature Pyramid Networks

• Original Paper

Mask R-CNN

• Original Paper

PointRend: Image Segmentation as Rendering

• Original Paper

---

Representation Learning

Unsupervised Feature Learning via Non-Parametric Instance Discrimination

• Original Paper

Momentum Contrast for Unsupervised Visual Representation Learning.

• Original Paper

A Simple Framework for Contrastive Learning of Visual Representations

• Original Paper

Bootstrap Your Own Latent- A New Approach to Self-supervised Learning

• Original Paper

Exploring Simple Siamese Representation Learning

• Original Paper

---

Generative Model

Generative Adversarial Networks

• Original Paper

A Style-Based Generator Architecture for Generative Adversarial Networks

• Original Paper

GAN Dissection: Visualizing and Understanding Generative Adversarial Networks

• Original Paper

Semantic Image Synthesis with Spatially-Adaptive Normalization

• Original Paper

Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks

• Original Paper

---

Image to Image Translation and Style Transfer

Image-to-Image Translation with Conditional Adversarial Networks (CVPR 2017)

• Original Paper

Bi-level Feature Alignment for Versatile Image Translation and Manipulation

• Original Paper

Instance Normalization: The Missing Ingredient for Fast Stylization

• Original Paper

Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization

• Original Paper

Semantic Image Synthesis with Spatially-Adaptive Normalization

• Original Paper

Universal Style Transfer via Feature Transforms

• Original Paper

A Neural Algorithm of Artistic Style

• Original Paper

---