Some interesting GitHub repositories
This is about the source code of Verilog language. It is a completely professional page, which is downloaded and inspected by the a lot of programmers. It includes every part of we need to do. (Contributor: Mehmet Saraçoğlu)
A collection of high-impact and state-of-the-art methods for superresolution problems using deep learning
Superresolution is a very interesting hot topic in deep learning. It involves inferring a higher-resolution image from a lower-resolution without losing the visual semantics of the image. This repository is a survey of the latest methods used in superresolution problems.
Contributor: Ege Can Kaya
V is a new programming language that is a compiled very fast, and provides several major improvements. Github page consists of the necessary codes about its compiler. Its page is created well and is organized. (Contributor: Mehmet Saraçoğlu)
PyTorch is a Python package for Tensor computation with strong GPU acceleration. It is widely used in deep learning problems and provides users with an excellent toolset suited for tensor manipulation and neural network creation.
Contributor: Ege Can Kaya
QT is a widely used GUI language. Well known programs like AutoCAD, KDE Display Manager, Telegram and more is created by this language. This repository is created nicely and holds information that we need to see. (Contributor: Mehmet Saraçoğlu)
CompressAI is a PyTorch library that provides easy to use data compression, specifically image/video compression, tools. It contains deep learning based data compression models that can be trained easily on many datasets. It also includes pre-trained state-of-the-art models for image compression.
- Has a detailed Wiki page that shows the installation and training processes. It also explains the structures of the models.
- Has evaluation scripts that measures the performance with respect to several metrics.
- Allows training directly on images, it converts images to tensors automatically.
- Models are written as Python classes. This allows users to modify and improve such models easily.
- Since learned image compression is a very recent research area, CompressAI is one of the very first efforts to assemble such research.
- Its license permits even commercial use in addition to private and patent use.
Contributor: Berkay Döner
This GitHub page includes some state-of-the-art face recognition models.
- Has a Wiki page that answers almost all errors. I only had to check Wiki for several errors I encountered.
- Serves many purposes such as identifying faces in a photo, drawing boxes around faces and blink detection.
- Very easy to use, it contains many examples for several cases that can be easily modified.
- Has a very high accuracy.
Contributor: Berkay Döner
This repository includes the source code of the Seaborn library of Python which is widely used in data science. It allows us quickly to create graphs, plots, and much more about a person needs while working a subject about data processing. (Contributor: Mehmet Saraçoğlu)
GTKWave is a program which shows the signals of a circuit. It waits the input in Verilog language, and shows a nice output using GUI. This repository is also created nicely and well-organized. (Contributor: Mehmet Saraçoğlu)
Flutters enables you to craft enjoyable experiences for users for mobile, web and desktop from a single codebase.
Some qualities of Flutter:
- Fully open source and they welcome contributors which makes this a community project.
- Works with existing code so you don't have to code it all over again.
- It's powered by the same hardware-accelerated 2D graphics library that underpins Chrome and Android: Skia. That means it's fast.
- With the help of stateful hot reload, Flutter let's you see the changes you made to your code instantly.
- Flutter works with whatever development tool you're using including Visual Studio Code.
This is a slime mold simulation. The name is based on Physarum polycephalum slime mold. This simple algorithm generates some cool outputs makes you feel as if you are looking through a microscope.
HuggingFace Transformers is one of the most powerful tools in modern NLP. It gathers implementations and pretrained weights of the state-of-the-art models into the single library by providing related interfaces for fast experimenting and fast deployment. The thing I most like about the Transformers is that it's a very useful tool for both research and productionizing the most up-to-date techniques. Also, the implementations provided inside the library are clean and very well-documented.
youtube-dl is the greatest video downloader for YouTube with a large number of options to customize. It can be used as either a standalone script or a Python package with the same functionality. It makes automated video downloading process super-easy for YouTube.
We can barely say that the Linux kernel is one of the most sophiscated open-source projects in the GitHub. The craziest thing about this repository is you can see the progress of the Linux kernel commit by commit since 2005. If you be a nice developer, you can see a recently committed code by Linus Torvalds himself.
AllenNLP is a well-designed NLP framework built on PyTorch for making development of new state-of-the-art models easier. It has a very flexible and easy-to-use structure. You can easily design your own brand-new architecture by using the default building blocks. Moreover, you can easily define new blocks to build your architecture with.
It is very big Github repository with 322k stars. Developers help each other and improve other developers software. Their curriculum is completely free and they organize many interactive coding challenges. It is very helpful to people who wants to learn coding for free. It also provides practices to new developers and very long tutorials for many software language in their youtube channel without any advertisement.
Contributor: İbrahim Kağan Bayat
John Washam was a developer who wants to be hired as a software engineer in Google and he made a to do list and uploaded it to the Github. It is very helpful for people who wants to be hired by a big company. It has many links to practicing websites, recommended books and many other useful things.
Contributor: İbrahim Kağan Bayat
Original Apollo 11 guidance computer source code for Command Module and Lunar Module. Humans has landed first time to the moon surface with the Apollo 11 spaceflight.
Contributor: İbrahim Kağan Bayat
(Contributor: Bengisu Özaydın)
- A source separation library with pretrained models.
- Provides pre-trained state-of-the-art models that have also been used by professional audio softwares.
- The model can separate an audio file into different stems. For example: 2-stem mode separates the file as vocals / accompaniment.
- Provides link to a Google Colab notebook that one can run online.
- I have been using this library and model to separate songs and practice singing over the accompaniment stem, as though I was doing karaoke.
- The library can be used to train other models.
(Contributor: Bengisu Özaydın)
- A curation of resources for software engineers.
- The most useful aspect of this repo is that quality is preferred over quantity in a curation.
- The resources are mostly technical, though important non-technical topics are also provided.
(Contributor: Bengisu Özaydın)
- The source code of the classical FPS shooter game, DOOM (1993), was made public in 1997.
- Includes a letter by the lead developer, John Carmack, that gives ideas about what can be done with the source code.
- Fun to tweak with. Can run on even many primitive architectures.
(Contributor: Bengisu Özaydın)
- A tamagotchi pet for the touchbar of Macbook Pro models.
- The touchbar pet needs attention, food and cleaning regularly.
- The pet and I keep each other company while I am working or studying intensively.
(Contributor: Bengisu Özaydın)
- A collection of useful .gitignore templates by GitHub.
- Amazing time-saver because the templates in this repo are enough for ordinary situations.
(Contributor: Bengisu Özaydın)
- A collection of tutorials on building many things from a search engine to an AR app.
- Good for me to see what excites me and what is boring to me.
(Contributor: Bengisu Özaydın)
- A collection of useful information and commands for the command line.
- Some parts are from answers to actual questions on Quora, which shows that it is a practical guide.
(Contributor: Bengisu Özaydın)
- This is a repo by the TAs of a lecture called "The Missing Semester" in MTI.
- The lecture covers many skills that a programmer had better know, but are not covered by the CS/CMPE curriculum.
(Contributor: Doğukan Akar)
Thingsboard is an IoT (Internet of Things) platform that allows to do almost anything for IoT. Using thingsboard you can manage your assets devices and customers and your devices can send the data they collected using thingsboard API. It help to visualize the data however you want. It mostly uses java and it uses javascript on the user interface. Things I liked in the Thingsboard are it is completely customizable and it is an open source platform that helps many IoT startups build their work quickly without worrying about UI and backend development.
(Contributor: Doğukan Akar)
Prophet is an R and Python module that helps doing forecasting and anomaly detection on time series data. It is built and maintained by the Facebook company. What I like on the Prophet is that it is very easy to learn and use, you just need to convert the data to the given shape. It helps both forecasting and plotting the results.