Open Source Objective C Artificial Intelligence Software

Flash-MoE is a high-performance implementation of mixture-of-experts (MoE) architectures designed to optimize the efficiency and scalability of large AI models. It focuses on accelerating routing and computation by leveraging optimized kernels and memory management techniques, allowing models to dynamically select specialized sub-networks during inference. The project aims to reduce the computational cost typically associated with MoE systems while maintaining or improving performance. It likely includes support for GPU acceleration and parallel processing, enabling it to handle large-scale workloads effectively. The architecture emphasizes speed and efficiency, making it suitable for both research and production environments where performance is critical. It may also provide tools for benchmarking and tuning model behavior. Overall, flash-moe represents a technical advancement in making MoE models more practical and deployable.

Image Crop Picker is an iOS/Android image picker with support for camera, video, configurable compression, multiple images and cropping. Module is creating tmp images which are going to be cleaned up automatically somewhere in the future. If you want to force cleanup, you can use clean to clean all tmp files, or cleanSingle(path) to clean single tmp file. Some of these types may not be available on all iOS versions.

Project Libre is an open platform for chat bots, virtual agents, live chat, and more. It includes web and mobile services for Java, Android, iOS, and JavaScript. Project Libre includes an API, SDK, and user interface for live chat, chatrooms, forums, chat bots, virtual agents, and other services supported by the Bot Libre and Paphus servers. The hosted servers include: BOT libre! - http://www.botlibre.com - free chat bot creation and hosting platform FORUMs libre! - http://www.forumslibre.com - free embeddable forums LIVE CHAT libre! - http://www.livechatlibre.com - free live chat and chatrooms Paphus Live Chat - http://www.paphuslivechat.com - commercial live chat, chatroom, forums, and chat bot hosting service Also hosted on GitHub, https://github.com/paphus/Project-Libre

A distributed capability manager for software modules and services. Components perform their specific tasks and may contribute information to an artificial intelligence engine, from which all components may benefit from the accumulated knowledge.

ASM is the Santa Fe Institute "Artificial Stock Market" simulation model. This page is mainly aimed to the revision/updating of ASM using the Swarm Toolkit, which is provided by Swarm Development Group (http://www.swarm.org).

ANE Training is an experimental research project that demonstrates how to train neural networks directly on Apple’s Neural Engine by leveraging reverse-engineered private APIs that are normally inaccessible to developers. The repository implements a from-scratch transformer training pipeline capable of running both forward and backward passes on ANE hardware without relying on CoreML, Metal, or GPU acceleration. It explores the internal software stack of the Apple Neural Engine by interfacing with private classes such as _ANEClient and compiling custom compute graphs in the MIL format. The project includes performance benchmarks and kernel breakdowns that show how different components of the training loop are distributed between the ANE and CPU. It is primarily intended as a research and educational proof of concept rather than a production library, highlighting what is technically possible with undocumented hardware access.

Computational modelling is a interdisciplinary field spanning many techniques and programming languages. AgenTools is a project dedicated to establishing increased modularity in scientific computing packages and a portal linking to other existing project

Quickly add voice to your app with the Alan Platform. Create an in-app voice assistant to enable human-like conversations and provide a personalized voice experience for every user. Alan is a conversational voice AI platform that lets you create an intelligent voice assistant for your app. It offers all the necessary tools to design, embed, and host your voice solutions. A powerful web-based IDE where you can write, test and debug dialog scenarios for your voice assistant or chatbot. Alan's AI-backend powered by the industry’s best Automatic Speech Recognition (ASR), Natural Language Understanding (NLU) and Speech Synthesis. The Alan Cloud provisions and handles the infrastructure required to maintain your voice deployments and perform all the voice processing tasks. Voice enable your app, you only need to get the Alan Client SDK and drop it into your app. No need to plan for, deploy and maintain any infrastructure or speech components - the Alan Platform does the bulk of the work.

BoomBot is a developer's framework for writing a bot connecting to the AIM network. It manages the network communication for the bots so that they can be easily written without worrying about AIM itself at all. Bots can be written in Objective-C or Lua.

Bot Libre is a free open source platform for artificial intelligence, chat bot, virtual agents, live chat, and more.

A class library written in Objective-C++ for building models of connectionist networks (neural networks). The class-hierarchy is designed to be maximally general and expandable, but also to support specific types of network.

"Distributed breve" (distbreve) is an open-source software package to make the process of implementing asynchronous, parallelizable steve code running under the breve simulation environment easily distributable amongst as many compute nodes as possible.

The Expedient Trickle Sync project is an attempt to create an intelligent algorithm for synchronizing data between a database server and a portable, mobile, network-enabled device, with the goal of minimizing the overall cost of data synchronization.

GLipKit (GL Image Processing Kit) is a collection of image processing and computer vision algorithms performed on current programmable 3D graphics hardware for faster execution. The main focus is currently computational stereo.

Implementation of a framework implementing various algorithms applying neural networks especially to digital signal processing. GNUstep is being used as the primary development environment.

Simulates a biologically diverse tropical island, and the ape inhabitants cognitive processes. For MacOS X, iPad, Windows and Linux. Features a non-polygonal graphics engine and a command-line version

Pangaea will be a robust and feature filled game engine built using Allegro (http://alleg.sourceforge.net) It will be similar to Final Fantasy 1-3, etc., complete with map editor/world builder.

Project Cogito is a artificial intelligence infrastructure to aid the development of thinking machines. It will provide the ability to create virtual worlds with loadable objects that brains can interact with. Administrative interfaces will be provided.

Prolog Image Processing is a (initially) MacOS based extension to OpenProlog which allows to use image processing function in the context of Prolog.

A speech-to-text library for React Native. Manually or automatically link the NativeModule. Drag the Voice.xcodeproj from the @react-native-voice/voice/ios folder to the Libraries group on Xcode in your project. Click on your main project file (the one that represents the .xcodeproj) select Build Phases and drag the static library, lib.Voice.a, from the Libraries/Voice.xcodeproj/Products folder to Link Binary With Libraries. The plugin provides props for extra customization. Every time you change the props or plugins, you'll need to rebuild (and prebuild) the native app. If no extra properties are added, defaults will be used. While the included VoiceTest app works without explicit permissions checks and requests, it may be necessary to add a permission request for RECORD_AUDIO for some configurations. Since Android M (6.0), user need to grant permission at runtime (and not during app installation).

RuleRule is a game which might be of interest to students and researchers of AI, machine learning and related areas. In this game, rules act and modify other rules (including themselves). The goal is to reach a desired condition (e.g. a checkerboard of red and yellow). Versions For a first test of the game mechanics in your browser, download the Javascript version, coded by Jan-Hendrik Vienup (December 2014, GPL) tested with Firefox, MSIE, Chrome, and Safari : RuleRule3.zip First created with NeXT Interface Builder and Objective-C files (1995, Mach OS): rulerule.gz (Creative Commons Attribution License). For some history, see http://hilgemeier.gmxhome.de/codes/rulerule.htm Unfinished Java project, running under Eclipse Luna with Windows 7 (2012, LGPL). Contains notes and design diagrams for possible future development. Doxygen-documented : RuleRule2.00_working.zip

SETI@home for Mac OS X. Now with style...

SPKML is a hybrid language being created initially as a web-based language that takes it's structure from existing language specifications, including: XML, XHTML, AIML (Artificial Intelligence Markup Language), and most importantly, PHP and all that works with HTML5 as well. The ability to be compiled into self executing cross-platorm bundles is part of the goal too...

The purpose of SNX is to have a modular series of components that report back to a central repository and 'AI' for pattern recognition and anomoly detection. This is to help with IDS systems, and to move to the next generation of security.

The unmanned autonomous vehicle is one step before the flying car. We propose a radically new approach which builds on the decades of research in Artificial Intelligence that is designed for a car from the ground up.