In this post, we’ve listed some awesome computer vision GitHub libraries that we hope will inspire all AI developers to build their projects just like this.

GitHub knowledge base for computer vision

1. 3D face reconstruction using CNN (★ — 4.1k)

This GitHub repository has a project where convolutional neural networks are used to reconstruct 3D face models using 2D images. This is a comprehensive repository, and we can choose to use the model in different languages (such as MATLAB, Python, etc.). To make it more interesting, we can even use our own images or our own examples and test them on this model.

2. Real-time multi-person pose estimation and Tracking system (★ — 4K)

This real-time multi-person pose estimation tracking system is called AlphaPose. The system is basically a way to map individual movements in real time. In addition, it can also estimate the individual’s posture. Again, this repository can help you gain more insight into how such a system works. This can also be a starting point for building software that takes advantage of this attitude estimation and tracking capability.

3, Using deep neural network for automatic coloring of photos (★ — 2.3K)

Another interesting computer vision project uses deep neural networks to color black and white photos. This computer vision GitHub repository contains Python code in the Jupyter notebook to make it easy to understand. In addition, there is a rich set of image data for training and testing models built for this task.

4, use generated neural network to edit nature photos (★ – 1.9k)

This repository is the code host for the research paper “Neural Photo Editing with Introspective advention Networks.” This project includes a simple interface that we can use to generate neural networks to edit nature photos. The current version is compatible with Python version 2.7, but there are still some inconsistencies with the latest version of Python.

5, Convolutional recursive Neural network for image recognition (★ — 1.7K)

This is a very interesting GitHub library where you can build an image recognition system using convolution recursive neural networks. The project is also useful in building scene text recognition and optical character recognition. The repository contains data sets for training and testing purposes, as well as demonstration examples.

6, Image removal blur using generated adversarial network (★ — 7.8k)

A lot of times we get annoyed with blurry images, and this GitHub repository has a solution. The PyTorch implementation of this paper entitled “De-blurring” basically takes a blurred image as input and uses a generation adversarial network to produce a clear image of the input. Again, the repository has complete source code and different kinds of data sets that can help you better understand and properly test the models you build.

7, Painting artificial intelligence — Deep reinforcement learning model, using strokes to generate painting (★ — 1.7K)

The Drawing AI GitHub library contains a model based on deep reinforcement learning that teaches machines to draw pictures of human drawings by using fewer strokes. Since it is based on reinforcement learning, the program does not require data for training purposes. Agents teach themselves to paint just like humans. I strongly encourage you to check out the warehouse and give it a try.

8. Lip Reading — Cross-audiovisual recognition using 3D architecture (★ — 1.4K)

Lip-reading is a computer vision project designed to solve problems encountered in audio and video streaming. The project uses audio-visual recognition to map audio and video. All of this is done using a 3D convolutional neural network architecture for mapping operations. The warehouse will certainly help build models to combat fake videos and other such misdeeds.

9, quick drawing – interactive drawing recognition tool (★ – 677)

Quickdraw is a computer vision project that can be used to identify a set of objects (similar objects) drawn using a pen. The model then tries to predict objects from a list of objects it has been trained to recognize. Quickdraw is basically an online game developed by Google. Another version of this project is to identify the project drawn on the canvas.

10, image animation using first-order motion model (★ — 3.9k)

This is a very amazing computer vision GitHub project where we can animate human faces from videos or images using our own faces as simulators. The model takes driving video and maps its motion onto still images to make the motion look real. The same concept applies to fashion data sets.

11, Fashion MNIST (★ — 7.8k)

This GitHub repository is made up of images of people wearing different kinds of clothes. The repository has a training set of 60,000 images and a test set of 10,000 images. Each image is a 28×28 grayscale image. It contains models built using available data sets. In general, the repository also helps validate your own machine learning algorithms by using it on your data sets. This is a beginner friendly data set, so they can get a feel for computer vision projects from this repository.

12, Cool Computer Vision Projects (★ — 37)

The library contains many interesting computer vision projects, such as face recognition, digital recognition, facial expression detection, object detection, object tracking, and so on. Through this library, you can learn some really cool things about computer vision. You can take inspiration from these projects or add features that extend them. This will really help you learn a lot and add to your personal project experience.

13, Intermediate Level Computer Vision Projects**(★ — 13) **

This is another useful GitHub repository that has multiple computer vision projects such as gesture recognition, face recognition, content-based image retrieval, and more. These are good projects for the intermediate level and will help improve your experience in the field of computer vision.

conclusion

In this article, we look at the GitHub repository for many computer vision projects to inspire you. We hope it will help you create your own computer vision projects, amaze others, and enhance your learning experience.