What are the most popular open source projects on GitHub and what new projects are crowding into the hot list? Take a look.
1, nocode
This is a newly released project in February. It can be said that it was the most popular Buddhist system programming in 2018. There is no one line of code in this project. The description of it is this: The best way to write secure and reliable applications. Write nothing; Nowhere. Some users have translated it into a Chinese version: To be safe and reliable, the best way is to deploy the entire world without a single line of code. This is an April Fool’s project, which has occupied the top spot for several days in the Github Trending List. There are now 16256 Stars.
2, tensorflow
TensorFlow is Google's second-generation machine learning system. According to Google, TensorFlow performed 2x faster than the first-generation DistBelief in some benchmark tests. TensorFlow has extensive support for built-in deep learning. Anything that can be expressed using computational flow graphs can use TensorFlow. Any gradient-based machine learning algorithm can benefit from the auto-differentiation of TensorFlow. With a flexible Python interface, it's easy to express ideas in TensorFlow.
3, front-end-interview-handbook
Different from other software engineers, the front-end job interviews pay less attention to the algorithm, and the knowledge in this field is more concentrated in Html, CSS, JavaScript and other aspects, although there are some resources on the Internet that can help front-end developers. Prepare for interviews, but they are not as informative as software engineer interviews. The front-end-interview-handbook is designed to prepare interview questions and answers for front-end job seekers.
4, content-aware image compression / expansion library caire
Caire is a content-aware image (size) compression/extension library based on the paper "Seam Carving for Content-Aware Image Resizing". The existing features are as follows:
Support custom command line
Supports reducing/enlargement of images
Adjust image size vertically/horizontally
Adjusts the size of all images in the directory
Does not require any third-party libraries
Fine tune with sobel thresholds
Use blur filters to enhance edge detection
Effect chart:
5, the data structure library Automerge
Automerge is a data structure library for building collaborative applications in JavaScript. A common way to build a JavaScript application is to save the state of the application in a model object, such as a JSON document. Has the following characteristics and design principles:
Network agnostics. Automerge is a pure data structure library that doesn't care what type of network you are using.
Unchanged state. An Automerge object is an immutable snapshot of the state of the application at some point in time. Whenever a change is made or a change from the network is merged, a new state object reflecting the change is returned.
Automatically merge. Automerge is a so-called collision-free replication data type (CRDT) that allows automatic consolidation of concurrent changes on different devices without requiring any central server.
Quite convenient. Automerge has been tested on Node.js, Chrome, Firefox and Electron.
6, developer-roadmap
This is a roadmap for a group of web developers built using Balsamiq, such as the front-end roadmap as follows:
Covers all important and hot topics such as front-end popular frameworks and languages.
7, FastPhotoStyle
FastPhotoStyle is an open source photorealistic transfer deep learning algorithm implementation library for Nvidia. Given a target photo and style reference photo, the code can convert the style photo style to a target photo to generate a new stylized photo.
The FastPhotoStyle implementation process consists of two steps: "stylized" and "fine tuned." First transfer the pattern of the style reference photo source to the target photo, and then fine tune it.
Effect chart:
8, Back-End-Developer-Interview-Questions
The previously mentioned front-end interviewing related open source projects, and this one is the back-end developers interviewing the whole, most of the problems covered are open-ended, some of which even have no answers, but they are precisely the most reflected Some problems with personal ability. At present, the project's questions are divided into 18 types, including code design issues, database issues, code version issues, distributed system issues, etc.
9, Java Component Services Framework Carbon
Carbon is a lightweight component model and enterprise-level service framework. Carbon is built on the following principles: Creating a more flexible framework A separate, metadata-centric system encourages separation of relationships, making code easier to maintain and reuse in a complex, dynamic operating environment Wait.
10, awesome-scalability
The project provides a large selection of reading materials designed to help back-end developers use the ideas of the various documents in the project to create a back-end with scalability, availability, and stability. Although this concept is vague, it is believed that we can understand and learn useful from the interpretation of famous engineers (Martin Fowler, Robert C. Martin, Tom White, etc.) and high quality resources (highscalability.com, infoq.com, etc.). thing.
The project covers all aspects. Mainly divided into principles, scalability, stability, other aspects, talks, books seven sections. Including exception handling, database strategies, and more.
11, android-ktx
Ndroid KTX is a set of extensions that make the Kotlin code on Android more concise, thereby improving the developer's programming experience. Parts of Android KTX that support the Android framework can now be found in the GitHub repository. What exactly does Android KTX do?
such as:
Strings are converted to URIs, and Android KTX adds an extension function to the string to make the strings more naturally converted to URIs.
Editing SharedPreferences, using Android KTX, the code is slightly shorter and can be read and written more naturally.
Translation path difference
Action on view onPreDraw
12, deepfake deep learning technology Facewap
Deepfake's deep learning technology, this tool's original purpose is to identify and exchange images and videos of people's facial images. The project has multiple entrances and you need to do:
Collect photos (or use the photos provided in the training data below)
Extract facial images from the original photos
Train the model on the photo (or use the model provided in the training data below)
Use model transformation source code
The 1080P high-definition projectors on the market can be roughly divided into three grades: high, medium, and low. The low-end products are mainly aimed at the entry-level market. In terms of performance, they can fully meet the needs of watching various HD videos. In terms of adjustment, it tends to be simpler, mainly to allow entry-level users to quickly set up the projector, but the contrast of such projectors is generally not very high, and the overall picture effect is relatively general; mid-range projectors mainly have certain projector knowledge. , the projection picture also has a better performance, the most important contrast of the cinema projector has also been greatly improved, the projector adjustment tends to be more professional, and it is also equipped with various additional functions to improve the projection effect; high-end products are Mainly aimed at enthusiast players, both price and performance are the symbol of flagship. For commercial use, in addition to satisfying home entertainment applications, 1080P high-definition standard projectors can also support clear and delicate text and table content projection for business office applications that often conduct office text, PPT presentations, and video presentations. .
4k meeting room projector,video projector for meeting room,conference room led projector,lcd meeting room projector,business 1080p HD projector
Shenzhen Happybate Trading Co.,LTD , https://www.happybateprojectors.com