Recommend 10 highly acclaimed and unique open source artificial intelligence projects I believe that everyone has seen or used a lot of artificial intelligence projects, but most of them seem to be very "high", making people feel that they have to master them like the skills of Xi Tulong. In fact, there are a lot of projects about artificial intelligence that are very practical, and the use is very interesting. Here is a simple list of 10 unique open source artificial intelligence projects. 1.STYLE2PAINTS: A powerful AI for coloring lines Https:// Please enter the title bcdef Recommended reason: The new generation of powerful line drafting AI can color the line draft according to the custom color uploaded by the user. The project provides an online use website that is very convenient to use. Please enter the title abcdefg 2.SerpentAI: Teaching AI to play the game learning framework Https:// Please enter the title bcdef Recommended reason: SerpentAI is designed to provide a valuable tool for machine learning and AI research. But at the same time, it is also very interesting for lovers. Please enter the title abcdefg 3.Synaptic.js: Neural Network Library for Browsers Https:// Please enter the title bcdef Recommended reason: Synaptic.js is a JavaScript neural network library for node.js and browsers that can build and train essentially any type of first-order or even second-order neural networks. The project has built in 4 classic neural network algorithms: multilayer perceptrons, long-short term memory networks, Liquid State Machine, and Hopfield neural networks. With Synaptic.js, you can easily test and compare the performance of different architectures. Please enter the title abcdefg 4.Snake-AI: artificial intelligence of the Snake game Https:// Please enter the title bcdef Recommended reason: an artificial intelligence of a snake game written in C/C++. The shortest path, the longest path, and the artificial intelligence algorithm are used. The purpose of AI is to let the snake eat as much food as possible until it fills the entire map. Demo Please enter the title abcdefg 5.Uncaptcha Https:// Please enter the title bcdef Recommended reason: crack the AI ​​algorithm of the reCAPTCHA system. The unCAPTCHA algorithm defeated the Google reCAPTCHA system with a 85% success rate. It relies on audio captcha attack - using browser automation software to parse the necessary elements and identify the phone number, and programmatically pass these numbers, eventually successfully spoofing the target site. Please enter the title abcdefg 6.Sockeye: Neural Machine Translation Framework Https:// Please enter the title bcdef Recommended reason: Sockeye is a fast and extensible deep learning library based on Apache MXNet. The Sockeye codebase has a unique advantage from MXNet. For example, through the symbolic and imperative MXNet APIs, Sockeye combines a declarative and imperative programming style; it can also train models in parallel on multiple GPUs. Sockeye implements the best sequence-to-sequence model currently available on MXNet. It also provides appropriate default values ​​for all sequence-to-sequence model hyperparameters. For optimization, there is no need to worry about stopping criteria, metric tracking, or weight initialization. You can simply run the provided Training Command Line Interface (CLI) or change the underlying model architecture easily. Please enter the title abcdefg 7.PHP-ML: PHP machine learning library Https:// Please enter the title bcdef Recommended reason: We all know that Python or C++ provides more machine learning libraries, but most of them are more complicated, and configuration makes many newcomers feel a headache. PHP-ML This machine learning library does not have a particularly tall algorithm, but it has the most basic algorithms for machine learning, classification, etc. Small projects or small companies do some simple data analysis, prediction, etc. PHP-ML is a machine learning library written in PHP. It also includes algorithms, cross-validation, neural networks, preprocessing, feature extraction, and more. Please enter the title abcdefg 8.CycleGAN: Generate anti-network image processing tools Https:// Please enter the title bcdef Recommended reason: This tool is very powerful, not only can "restore" the painting into a photo (can be understood as a "filter"), but also convert summer into winter, or convert ordinary horse into zebra. Unlike other artificial intelligence paintings, CycleGAN's research team is trying to create a two-way algorithm that can convert information in both directions without losing information. The details of the photos in CycleGAN were required to be completely retained. The researchers hoped that after entering a picture into CycleGAN, they would perform multiple iterations (photo → painting → photo → painting → photo), and finally get the same or similar picture as the original photo. . Please enter the title abcdefg 9.DeepLearn.js: Machine Learning JS Library for Accelerating Hardware Https:// Please enter the title bcdef DeepLearn.js is an open source JavaScript library from Google that can be used for machine intelligence and accelerates WebGL. It runs entirely in the browser, does not require installation, and does not require backend processing. DeepLearn.js provides an efficient machine learning building block that allows us to train neural networks in the browser or run pre-trained models in inferential mode. It provides an API for building microdata graphs and a set of mathematical functions that can be used directly. Although machine learning libraries on browsers have existed for many years (such as Andrej Karpathy's convnetjs), they are limited by the speed of JavaScript, or limited to reasoning and not for training (such as TensorFire). In contrast, deeplearn.js achieves significant acceleration by leveraging WebGL to perform computations on the GPU and the ability to perform full backpropagation. Please enter the title abcdefg 10.TensorFire: Browser-side neural network framework Https:// Please enter the title bcdef Recommended reason: TensorFire is a WebGL-based, neural network framework that runs in a browser. Applications written with TensorFire can implement leading edge deep learning algorithms while running directly in modern browsers without any installation or configuration. Compared to previous neural network frameworks in some browsers, TensorFire has nearly a hundred times faster speeds, even comparable to those running on local CPUs. Developers can also use the underlying interfaces provided by TensorFire for other high-performance computing such as PageRank, cellular automaton emulation, image conversion and filtering, and more. Wuxi Lerin New Energy Technology Co.,Ltd. , https://www.lerin-tech.com
