[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-onmyway133--awesome-machine-learning":3,"tool-onmyway133--awesome-machine-learning":61},[4,18,26,36,44,53],{"id":5,"name":6,"github_repo":7,"description_zh":8,"stars":9,"difficulty_score":10,"last_commit_at":11,"category_tags":12,"status":17},4358,"openclaw","openclaw\u002Fopenclaw","OpenClaw 是一款专为个人打造的本地化 AI 助手，旨在让你在自己的设备上拥有完全可控的智能伙伴。它打破了传统 AI 助手局限于特定网页或应用的束缚，能够直接接入你日常使用的各类通讯渠道，包括微信、WhatsApp、Telegram、Discord、iMessage 等数十种平台。无论你在哪个聊天软件中发送消息，OpenClaw 都能即时响应，甚至支持在 macOS、iOS 和 Android 设备上进行语音交互，并提供实时的画布渲染功能供你操控。\n\n这款工具主要解决了用户对数据隐私、响应速度以及“始终在线”体验的需求。通过将 AI 部署在本地，用户无需依赖云端服务即可享受快速、私密的智能辅助，真正实现了“你的数据，你做主”。其独特的技术亮点在于强大的网关架构，将控制平面与核心助手分离，确保跨平台通信的流畅性与扩展性。\n\nOpenClaw 非常适合希望构建个性化工作流的技术爱好者、开发者，以及注重隐私保护且不愿被单一生态绑定的普通用户。只要具备基础的终端操作能力（支持 macOS、Linux 及 Windows WSL2），即可通过简单的命令行引导完成部署。如果你渴望拥有一个懂你",349277,3,"2026-04-06T06:32:30",[13,14,15,16],"Agent","开发框架","图像","数据工具","ready",{"id":19,"name":20,"github_repo":21,"description_zh":22,"stars":23,"difficulty_score":10,"last_commit_at":24,"category_tags":25,"status":17},3808,"stable-diffusion-webui","AUTOMATIC1111\u002Fstable-diffusion-webui","stable-diffusion-webui 是一个基于 Gradio 构建的网页版操作界面，旨在让用户能够轻松地在本地运行和使用强大的 Stable Diffusion 图像生成模型。它解决了原始模型依赖命令行、操作门槛高且功能分散的痛点，将复杂的 AI 绘图流程整合进一个直观易用的图形化平台。\n\n无论是希望快速上手的普通创作者、需要精细控制画面细节的设计师，还是想要深入探索模型潜力的开发者与研究人员，都能从中获益。其核心亮点在于极高的功能丰富度：不仅支持文生图、图生图、局部重绘（Inpainting）和外绘（Outpainting）等基础模式，还独创了注意力机制调整、提示词矩阵、负向提示词以及“高清修复”等高级功能。此外，它内置了 GFPGAN 和 CodeFormer 等人脸修复工具，支持多种神经网络放大算法，并允许用户通过插件系统无限扩展能力。即使是显存有限的设备，stable-diffusion-webui 也提供了相应的优化选项，让高质量的 AI 艺术创作变得触手可及。",162132,"2026-04-05T11:01:52",[14,15,13],{"id":27,"name":28,"github_repo":29,"description_zh":30,"stars":31,"difficulty_score":32,"last_commit_at":33,"category_tags":34,"status":17},1381,"everything-claude-code","affaan-m\u002Feverything-claude-code","everything-claude-code 是一套专为 AI 编程助手（如 Claude Code、Codex、Cursor 等）打造的高性能优化系统。它不仅仅是一组配置文件，而是一个经过长期实战打磨的完整框架，旨在解决 AI 代理在实际开发中面临的效率低下、记忆丢失、安全隐患及缺乏持续学习能力等核心痛点。\n\n通过引入技能模块化、直觉增强、记忆持久化机制以及内置的安全扫描功能，everything-claude-code 能显著提升 AI 在复杂任务中的表现，帮助开发者构建更稳定、更智能的生产级 AI 代理。其独特的“研究优先”开发理念和针对 Token 消耗的优化策略，使得模型响应更快、成本更低，同时有效防御潜在的攻击向量。\n\n这套工具特别适合软件开发者、AI 研究人员以及希望深度定制 AI 工作流的技术团队使用。无论您是在构建大型代码库，还是需要 AI 协助进行安全审计与自动化测试，everything-claude-code 都能提供强大的底层支持。作为一个曾荣获 Anthropic 黑客大奖的开源项目，它融合了多语言支持与丰富的实战钩子（hooks），让 AI 真正成长为懂上",154349,2,"2026-04-13T23:32:16",[14,13,35],"语言模型",{"id":37,"name":38,"github_repo":39,"description_zh":40,"stars":41,"difficulty_score":32,"last_commit_at":42,"category_tags":43,"status":17},2271,"ComfyUI","Comfy-Org\u002FComfyUI","ComfyUI 是一款功能强大且高度模块化的视觉 AI 引擎，专为设计和执行复杂的 Stable Diffusion 图像生成流程而打造。它摒弃了传统的代码编写模式，采用直观的节点式流程图界面，让用户通过连接不同的功能模块即可构建个性化的生成管线。\n\n这一设计巧妙解决了高级 AI 绘图工作流配置复杂、灵活性不足的痛点。用户无需具备编程背景，也能自由组合模型、调整参数并实时预览效果，轻松实现从基础文生图到多步骤高清修复等各类复杂任务。ComfyUI 拥有极佳的兼容性，不仅支持 Windows、macOS 和 Linux 全平台，还广泛适配 NVIDIA、AMD、Intel 及苹果 Silicon 等多种硬件架构，并率先支持 SDXL、Flux、SD3 等前沿模型。\n\n无论是希望深入探索算法潜力的研究人员和开发者，还是追求极致创作自由度的设计师与资深 AI 绘画爱好者，ComfyUI 都能提供强大的支持。其独特的模块化架构允许社区不断扩展新功能，使其成为当前最灵活、生态最丰富的开源扩散模型工具之一，帮助用户将创意高效转化为现实。",108322,"2026-04-10T11:39:34",[14,15,13],{"id":45,"name":46,"github_repo":47,"description_zh":48,"stars":49,"difficulty_score":32,"last_commit_at":50,"category_tags":51,"status":17},6121,"gemini-cli","google-gemini\u002Fgemini-cli","gemini-cli 是一款由谷歌推出的开源 AI 命令行工具，它将强大的 Gemini 大模型能力直接集成到用户的终端环境中。对于习惯在命令行工作的开发者而言，它提供了一条从输入提示词到获取模型响应的最短路径，无需切换窗口即可享受智能辅助。\n\n这款工具主要解决了开发过程中频繁上下文切换的痛点，让用户能在熟悉的终端界面内直接完成代码理解、生成、调试以及自动化运维任务。无论是查询大型代码库、根据草图生成应用，还是执行复杂的 Git 操作，gemini-cli 都能通过自然语言指令高效处理。\n\n它特别适合广大软件工程师、DevOps 人员及技术研究人员使用。其核心亮点包括支持高达 100 万 token 的超长上下文窗口，具备出色的逻辑推理能力；内置 Google 搜索、文件操作及 Shell 命令执行等实用工具；更独特的是，它支持 MCP（模型上下文协议），允许用户灵活扩展自定义集成，连接如图像生成等外部能力。此外，个人谷歌账号即可享受免费的额度支持，且项目基于 Apache 2.0 协议完全开源，是提升终端工作效率的理想助手。",100752,"2026-04-10T01:20:03",[52,13,15,14],"插件",{"id":54,"name":55,"github_repo":56,"description_zh":57,"stars":58,"difficulty_score":32,"last_commit_at":59,"category_tags":60,"status":17},4721,"markitdown","microsoft\u002Fmarkitdown","MarkItDown 是一款由微软 AutoGen 团队打造的轻量级 Python 工具，专为将各类文件高效转换为 Markdown 格式而设计。它支持 PDF、Word、Excel、PPT、图片（含 OCR）、音频（含语音转录）、HTML 乃至 YouTube 链接等多种格式的解析，能够精准提取文档中的标题、列表、表格和链接等关键结构信息。\n\n在人工智能应用日益普及的今天，大语言模型（LLM）虽擅长处理文本，却难以直接读取复杂的二进制办公文档。MarkItDown 恰好解决了这一痛点，它将非结构化或半结构化的文件转化为模型“原生理解”且 Token 效率极高的 Markdown 格式，成为连接本地文件与 AI 分析 pipeline 的理想桥梁。此外，它还提供了 MCP（模型上下文协议）服务器，可无缝集成到 Claude Desktop 等 LLM 应用中。\n\n这款工具特别适合开发者、数据科学家及 AI 研究人员使用，尤其是那些需要构建文档检索增强生成（RAG）系统、进行批量文本分析或希望让 AI 助手直接“阅读”本地文件的用户。虽然生成的内容也具备一定可读性，但其核心优势在于为机器",93400,"2026-04-06T19:52:38",[52,14],{"id":62,"github_repo":63,"name":64,"description_en":65,"description_zh":66,"ai_summary_zh":67,"readme_en":68,"readme_zh":69,"quickstart_zh":70,"use_case_zh":71,"hero_image_url":72,"owner_login":73,"owner_name":74,"owner_avatar_url":75,"owner_bio":76,"owner_company":77,"owner_location":78,"owner_email":79,"owner_twitter":73,"owner_website":80,"owner_url":81,"languages":82,"stars":83,"forks":84,"last_commit_at":85,"license":86,"difficulty_score":87,"env_os":88,"env_gpu":89,"env_ram":89,"env_deps":90,"category_tags":103,"github_topics":104,"view_count":32,"oss_zip_url":82,"oss_zip_packed_at":82,"status":17,"created_at":115,"updated_at":116,"faqs":117,"releases":118},7379,"onmyway133\u002Fawesome-machine-learning","awesome-machine-learning","🎰 A curated list of machine learning resources, preferably CoreML","awesome-machine-learning 是一个精心整理的机器学习资源清单，特别聚焦于苹果生态下的 Core ML 框架与 Swift 语言。它旨在解决开发者在尝试机器学习时，无需深入 Python 或 JavaScript 等其他平台即可快速上手的痛点。通过汇集预训练模型、转换工具及教程，它让使用者能先直观感受技术成果，再逐步探索底层机制。\n\n这份资源非常适合 iOS 和 macOS 应用开发者，尤其是那些希望利用 Apple 原生技术（如 Vision、ARKit）在移动端集成智能功能，但不想被复杂跨平台环境困扰的人群。同时，对希望将现有 TensorFlow、Keras 或 Caffe 模型迁移至苹果设备的算法工程师也极具参考价值。\n\n其独特亮点在于提供了丰富的模型转换工具链（如 tf-coreml、torch2coreml），支持将主流深度学习框架模型无缝转为 Core ML 格式，并收录了专为 iOS 优化的模型库（如 Awesome-CoreML-Models）。此外，清单还涵盖了从模型可视化查看器 Netron 到风格迁移实战项目 StyleArt 等实用内容，帮","awesome-machine-learning 是一个精心整理的机器学习资源清单，特别聚焦于苹果生态下的 Core ML 框架与 Swift 语言。它旨在解决开发者在尝试机器学习时，无需深入 Python 或 JavaScript 等其他平台即可快速上手的痛点。通过汇集预训练模型、转换工具及教程，它让使用者能先直观感受技术成果，再逐步探索底层机制。\n\n这份资源非常适合 iOS 和 macOS 应用开发者，尤其是那些希望利用 Apple 原生技术（如 Vision、ARKit）在移动端集成智能功能，但不想被复杂跨平台环境困扰的人群。同时，对希望将现有 TensorFlow、Keras 或 Caffe 模型迁移至苹果设备的算法工程师也极具参考价值。\n\n其独特亮点在于提供了丰富的模型转换工具链（如 tf-coreml、torch2coreml），支持将主流深度学习框架模型无缝转为 Core ML 格式，并收录了专为 iOS 优化的模型库（如 Awesome-CoreML-Models）。此外，清单还涵盖了从模型可视化查看器 Netron 到风格迁移实战项目 StyleArt 等实用内容，帮助用户高效构建具备图像识别、自然语言处理等能力的原生智能应用。","# awesome-machine-learning [![Awesome](https:\u002F\u002Fcdn.rawgit.com\u002Fsindresorhus\u002Fawesome\u002Fd7305f38d29fed78fa85652e3a63e154dd8e8829\u002Fmedia\u002Fbadge.svg)](https:\u002F\u002Fgithub.com\u002Fsindresorhus\u002Fawesome)\n\n❤️ Support my apps ❤️ \n\n- [Push Hero - pure Swift native macOS application to test push notifications](https:\u002F\u002Fonmyway133.com\u002Fpushhero)\n- [PastePal - Pasteboard, note and shortcut manager](https:\u002F\u002Fonmyway133.com\u002Fpastepal)\n- [Quick Check - smart todo manager](https:\u002F\u002Fonmyway133.com\u002Fquickcheck)\n- [Alias - App and file shortcut manager](https:\u002F\u002Fonmyway133.com\u002Falias)\n- [My other apps](https:\u002F\u002Fonmyway133.com\u002Fapps\u002F)\n\n❤️❤️😇😍🤘❤️❤️\n\nI like to explore machine learning, but don't want the to dive into other platforms, like Python or Javascript, to understand some frameworks, or TensorFlow. Luckily, at WWDC 2017, Apple introduces Core ML, Vision, ARKit, which makes working with machine learning so much easier. With all the pre-trained models, we can build great things. It's good to feel the outcome first, then try to explore advanced topics and underlying mechanisms 🤖\n\nThis will curates things mostly related to Core ML, and Swift. There are related things in other platforms if you want to get some references\n\n## Table of contents\n\n- [Core ML](#core-ml)\n- [TensorFlow](#tensorflow)\n- [Keras](#keras)\n- [Turi Create](#turi-create)\n- [Machine Learning](#machine-learning)\n- [Misc](#misc)\n\n## Core ML\n\n### Models :rocket:\n\n- [Awesome-CoreML-Models](https:\u002F\u002Fgithub.com\u002Flikedan\u002FAwesome-CoreML-Models) Largest list of models for Core ML (for iOS 11+)\n- [caffe](https:\u002F\u002Fgithub.com\u002FBVLC\u002Fcaffe) Caffe: a fast open framework for deep learning. http:\u002F\u002Fcaffe.berkeleyvision.org\u002F\n- [deep-learning-models](https:\u002F\u002Fgithub.com\u002Ffchollet\u002Fdeep-learning-models) Keras code and weights files for popular deep learning models.\n- [tensorflow models](https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels) Models built with TensorFlow\n- [libSVM](https:\u002F\u002Fwww.csie.ntu.edu.tw\u002F~cjlin\u002Flibsvm\u002F) A Library for Support Vector Machines\n- [scikit-learn](http:\u002F\u002Fscikit-learn.org\u002F) Machine Learning in Python\n- [xgboost](https:\u002F\u002Fgithub.com\u002Fdmlc\u002Fxgboost) Scalable, Portable and Distributed Gradient Boosting (GBDT, GBRT or GBM) Library, for Python, R, Java, Scala, C++ and more. Runs on single machine, Hadoop, Spark, Flink and DataFlow\n- [Keras-Classification-Models](https:\u002F\u002Fgithub.com\u002Ftitu1994\u002FKeras-Classification-Models) Collection of Keras models used for classification\n- [MobileNet-Caffe](https:\u002F\u002Fgithub.com\u002Fshicai\u002FMobileNet-Caffe) Caffe Implementation of Google's MobileNets\n- [ModelZoo](https:\u002F\u002Fgithub.com\u002Fcocoa-ai\u002FModelZoo) A central GitHub repository for sharing Core ML models \n- [StyleArt](https:\u002F\u002Fgithub.com\u002Fileafsolutions\u002FStyleArt) Style Art library process images using COREML with a set of pre trained machine learning models and convert them to Art style\n- [models](https:\u002F\u002Fgithub.com\u002FSarasra\u002Fmodels) Models and examples built with TensorFlow\n- [Core ML Store](https:\u002F\u002Fcoreml.store\u002F)\n\n### Tools\n\n- [coremltools](https:\u002F\u002Fpypi.python.org\u002Fpypi\u002Fcoremltools) coremltools in a python package for creating, examining, and testing models in the .mlmodel format\n- [torch2coreml](https:\u002F\u002Fgithub.com\u002Fprisma-ai\u002Ftorch2coreml) This tool helps convert Torch7 models into Apple CoreML \n- [turicreate](https:\u002F\u002Fgithub.com\u002Fapple\u002Fturicreate) Turi Create simplifies the development of custom machine learning models.\n- [Netron](https:\u002F\u002Fgithub.com\u002Flutzroeder\u002FNetron) Viewer for neural network and machine learning models\n- [onnx-coreml](https:\u002F\u002Fgithub.com\u002Fonnx\u002Fonnx-coreml) ONNX to CoreML Converter\n- [tf-coreml](https:\u002F\u002Fgithub.com\u002Ftf-coreml\u002Ftf-coreml) TensorFlow to CoreML Converter\n- [tensorwatch](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Ftensorwatch) Debugging and visualization tool for machine learning and data science\n\n### Posts\n\n- [Swift Tutorial: Native Machine Learning and Machine Vision in iOS 11](https:\u002F\u002Fhackernoon.com\u002Fswift-tutorial-native-machine-learning-and-machine-vision-in-ios-11-11e1e88aa397)\n- [How to train your own model for CoreML](http:\u002F\u002Freza.codes\u002F2017-07-29\u002Fhow-to-train-your-own-dataset-for-coreml\u002F)\n- [Smart Gesture Recognition in iOS 11 with Core ML and TensorFlow](https:\u002F\u002Fhackernoon.com\u002Fsmart-gesture-recognition-in-ios-11-with-core-ml-and-tensorflow-1a0a92c99c51)\n- [DIY Prisma app with CoreML](https:\u002F\u002Fblog.prismalabs.ai\u002Fdiy-prisma-app-with-coreml-6b4994cc99e1) :rocket:\n- [Using scikit-learn and CoreML to Create a Music Recommendation Engine](https:\u002F\u002Fwww.agnosticdev.com\u002Fblog-entry\u002Fpython\u002Fusing-scikit-learn-and-coreml-create-music-recommendation-engine)\n- [Building Not Hotdog with Turi Create and Core ML — in an afternoon](https:\u002F\u002Fheartbeat.fritz.ai\u002Fbuilding-not-hotdog-with-turi-create-and-core-ml-in-an-afternoon-a87fd1967d10)\n- [Build a Taylor Swift detector with the TensorFlow Object Detection API, ML Engine, and Swift](https:\u002F\u002Ftowardsdatascience.com\u002Fbuild-a-taylor-swift-detector-with-the-tensorflow-object-detection-api-ml-engine-and-swift-82707f5b4a56)\n- [Leveraging Machine Learning in iOS For Improved Accessibility](https:\u002F\u002Fmedium.com\u002Fbuffer-engineering\u002Fleveraging-machine-learning-in-ios-for-improved-accessibility-fc7796c5326f)\n- [IBM Watson Services for Core ML Tutorial](https:\u002F\u002Fwww.raywenderlich.com\u002F190191\u002Fibm-watson-services-for-core-ml-tutorial)\n- [Beginning Machine Learning with Keras & Core ML](https:\u002F\u002Fwww.raywenderlich.com\u002F181760\u002Fbeginning-machine-learning-keras-core-ml)\n- [Detecting Whisky brands with Core ML and IBM Watson services](https:\u002F\u002Fmartinmitrevski.com\u002F2018\u002F04\u002F14\u002Fdetecting-whisky-brands-with-core-ml-and-ibm-watson-services\u002F)\n- [Detecting Avengers superheroes in your iOS app with IBM Watson and CoreML](https:\u002F\u002Fmedium.com\u002Fflawless-app-stories\u002Fdetecting-avengers-superheroes-in-your-ios-app-with-ibm-watson-and-coreml-fe38e493a4d1)\n- [Machine Learning](https:\u002F\u002Fdeveloper.apple.com\u002Fmachine-learning\u002F) Build more intelligent apps with machine learning.\n- [Apple Machine Learning Journal](https:\u002F\u002Fmachinelearning.apple.com\u002F)\n- [Introducing Core ML](https:\u002F\u002Fdeveloper.apple.com\u002Fvideos\u002Fplay\u002Fwwdc2017\u002F703\u002F)\n- [Core ML in depth](https:\u002F\u002Fdeveloper.apple.com\u002Fvideos\u002Fplay\u002Fwwdc2017\u002F710\u002F)\n- [Core ML and Vision: Machine Learning in iOS 11 Tutorial](https:\u002F\u002Fwww.raywenderlich.com\u002F164213\u002Fcoreml-and-vision-machine-learning-in-ios-11-tutorial)\n- [iOS 11: Machine Learning for everyone](http:\u002F\u002Fmachinethink.net\u002Fblog\u002Fios-11-machine-learning-for-everyone\u002F)\n- [Everything a Swift Dev Ever Wanted to Know About Machine Learning](https:\u002F\u002Fnews.realm.io\u002Fnews\u002Fswift-developer-on-machine-learning-try-swift-2017-gallagher\u002F)\n- [Bringing machine learning to your iOS apps](https:\u002F\u002Fnews.realm.io\u002Fnews\u002Faltconf-2017-meghan-kane-bringing-machine-learning-to-your-ios-apps\u002F)\n- [Pros and cons of iOS machine learning APIs](http:\u002F\u002Fmachinethink.net\u002Fblog\u002Fmachine-learning-apis\u002F) :rocket:\n- [Core ML: Machine Learning for iOS](https:\u002F\u002Fwww.udacity.com\u002Fcourse\u002Fcore-ml--ud1038) :rocket:\n- [Bootstrapping the Machine Learning Training Process](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=ugiPfm8ICZo)\n- [Detecting Pneumonia in an iOS App with Create ML](https:\u002F\u002Fheartbeat.fritz.ai\u002Fdetecting-pneumonia-in-an-ios-app-with-create-ml-5cff2a60a3d)\n- [How to fine-tune ResNet in Keras and use it in an iOS App via Core ML](https:\u002F\u002Fheartbeat.fritz.ai\u002Fhow-to-fine-tune-resnet-in-keras-and-use-it-in-an-ios-app-via-core-ml-ee7fd84c1b26)\n- [Five Common Data Quality Gotchas in Machine Learning and How to Detect Them Quickly](https:\u002F\u002Fdoordash.engineering\u002F2022\u002F09\u002F27\u002Ffive-common-data-quality-gotchas-in-machine-learning-and-how-to-detect-them-quickly\u002F)\n\n### Repos\n\n- [Core-ML-Sample](https:\u002F\u002Fgithub.com\u002Fyulingtianxia\u002FCore-ML-Sample) A Demo using Core ML Framework\n- [UnsplashExplorer-CoreML](https:\u002F\u002Fgithub.com\u002Fahmetws\u002FUnsplashExplorer-CoreML) Core ML demo app with Unsplash API\n- [MNIST_DRAW](https:\u002F\u002Fgithub.com\u002Fhwchong\u002FMNIST_DRAW) This is a sample project demonstrating the use of Keras (Tensorflow) for the training of a MNIST model for handwriting recognition using CoreML on iOS 11 for inference.\n- [CocoaAI](https:\u002F\u002Fgithub.com\u002Fcocoa-ai\u002FCocoaAI) The Cocoa Artificial Intelligence Lab :rocket:\n- [complex-gestures-demo](https:\u002F\u002Fgithub.com\u002Fmitochrome\u002Fcomplex-gestures-demo) A demonstration of using machine learning to recognize 13 complex gestures in an iOS app\n - [Core-ML-Car-Recognition](https:\u002F\u002Fgithub.com\u002Flikedan\u002FCore-ML-Car-Recognition) A Car Recognition Framework for CoreML\n - [CoreML-in-ARKit](https:\u002F\u002Fgithub.com\u002Fhanleyweng\u002FCoreML-in-ARKit) Simple project to detect objects and display 3D labels above them in AR. This serves as a basic template for an ARKit project to use CoreML\n - [trainer-mac](https:\u002F\u002Fgithub.com\u002Fmortenjust\u002Ftrainer-mac) Trains a model, then generates a complete Xcode project that uses it - no code necessary\n - [GestureAI-CoreML-iOS](https:\u002F\u002Fgithub.com\u002Fakimach\u002FGestureAI-CoreML-iOS) Hand-gesture recognition on iOS app using CoreML\n - [visual-recognition-coreml](https:\u002F\u002Fgithub.com\u002Fwatson-developer-cloud\u002Fvisual-recognition-coreml) Classify images offline using Watson Visual Recognition and Core ML\n \n## TensorFlow\n\n### Posts\n\n- [Open Source TensorFlow Models (Google I\u002FO '17)](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=9ziVGkt8Gg4)\n- [Swift for TensorFlow](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=Yze693W4MaU)\n- [Get started with TensorFlow high-level APIs (Google I\u002FO '18)](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=tjsHSIG8I08&list=PLOU2XLYxmsIInFRc3M44HUTQc3b_YJ4-Y&index=37)\n- [Getting started with TensorFlow on iOS](http:\u002F\u002Fmachinethink.net\u002Fblog\u002Ftensorflow-on-ios\u002F)\n- [Introducing TensorFlow.js: Machine Learning in Javascript](https:\u002F\u002Fmedium.com\u002Ftensorflow\u002Fintroducing-tensorflow-js-machine-learning-in-javascript-bf3eab376db)\n- [TensorFlow for JavaScript (Google I\u002FO '18)](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=OmofOvMApTU)\n- [Colab](https:\u002F\u002Fcolab.research.google.com) Colaboratory is a Google research project created to help disseminate machine learning education and research\n- [Use TensorFlow and BNNS to Add Machine Learning to your Mac or iOS App](https:\u002F\u002Fwww.bignerdranch.com\u002Fblog\u002Fuse-tensorflow-and-bnns-to-add-machine-learning-to-your-mac-or-ios-app\u002F)\n\n### Repos\n\n- [workshops](https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fworkshops) A few exercises for use at events Google IO 2018\n\n### Courses\n\n- [Machine Learning Crash Course with TensorFlow APIs](https:\u002F\u002Fdevelopers.google.com\u002Fmachine-learning\u002Fcrash-course\u002F)\n- [Deep Learning with Python by Francois Chollet](https:\u002F\u002Fwww.amazon.com\u002FDeep-Learning-Python-Francois-Chollet\u002Fdp\u002F1617294438)\n- [Learn from ML experts at Google](https:\u002F\u002Fai.google\u002Feducation)\n\n## Keras\n\n### Posts\n\n- [Running Keras models on iOS with CoreML](https:\u002F\u002Fwww.pyimagesearch.com\u002F2018\u002F04\u002F23\u002Frunning-keras-models-on-ios-with-coreml\u002F)\n- [Keras and Convolutional Neural Networks (CNNs)](https:\u002F\u002Fwww.pyimagesearch.com\u002F2018\u002F04\u002F16\u002Fkeras-and-convolutional-neural-networks-cnns\u002F)\n- [Keras Tutorial : Transfer Learning using pre-trained models]()https:\u002F\u002Fwww.learnopencv.com\u002Fkeras-tutorial-transfer-learning-using-pre-trained-models\u002F\n\n## Turi Create\n\n### Posts\n\n- [Building Not Hotdog with Turi Create and Core ML — in an afternoon](https:\u002F\u002Fhackernoon.com\u002Fbuilding-not-hotdog-with-turi-create-and-core-ml-in-an-afternoon-231b14738edf)\n- [Natural Language Processing on iOS with Turi Create](https:\u002F\u002Fwww.raywenderlich.com\u002F185515\u002Fnatural-language-processing-on-ios-with-turi-create)\n- [Machine Learning in iOS: Turi Create and CoreML](https:\u002F\u002Fmedium.com\u002Fflawless-app-stories\u002Fmachine-learning-in-ios-turi-create-and-coreml-5ddce0dc8e26)\n- [A Guide to Turi Create](https:\u002F\u002Fdeveloper.apple.com\u002Fvideos\u002Fplay\u002Fwwdc2018\u002F712\u002F)\n\n## Machine Learning \n\n### Getting started\n\n- [A Thank You note to Towards Data Science](https:\u002F\u002Ftowardsdatascience.com\u002Fa-thank-you-note-to-towards-data-science-58b714a824f8)\n- [This is why anyone can learn Machine Learning](https:\u002F\u002Fmedium.freecodecamp.org\u002Fthis-is-why-anyone-can-learn-machine-learning-a5333ee64dff)\n\n### Posts\n\n- [A visual introduction to machine learning](http:\u002F\u002Fwww.r2d3.us\u002Fvisual-intro-to-machine-learning-part-1\u002F)\n- [Machine Learning is Fun!](https:\u002F\u002Fmedium.com\u002F@ageitgey\u002Fmachine-learning-is-fun-80ea3ec3c471)\n- [10 Machine Learning Terms Explained in Simple English](http:\u002F\u002Fblog.aylien.com\u002F10-machine-learning-terms-explained-in-simple\u002F)\n- [Machine Learning in a Year](https:\u002F\u002Fmedium.com\u002Flearning-new-stuff\u002Fmachine-learning-in-a-year-cdb0b0ebd29c)\n- [Machine Learning Self-study Resources](https:\u002F\u002Fragle.sanukcode.net\u002Farticles\u002Fmachine-learning-self-study-resources\u002F)\n- [How to Learn Machine Learning](https:\u002F\u002Felitedatascience.com\u002Flearn-machine-learning)\n- [Getting Started with Machine Learning](https:\u002F\u002Fmedium.com\u002F@suffiyanz\u002Fgetting-started-with-machine-learning-f15df1c283ea)\n- [The Non-Technical Guide to Machine Learning & Artificial Intelligence](https:\u002F\u002Fmachinelearnings.co\u002Fa-humans-guide-to-machine-learning-e179f43b67a0)\n- [Machine Learning: An In-Depth Guide - Overview, Goals, Learning Types, and Algorithms](http:\u002F\u002Fwww.innoarchitech.com\u002Fmachine-learning-an-in-depth-non-technical-guide\u002F)\n- [A Tour of Machine Learning Algorithms](http:\u002F\u002Fmachinelearningmastery.com\u002Fa-tour-of-machine-learning-algorithms\u002F)\n- [Machine Learning for Hackers](https:\u002F\u002Fwww.youtube.com\u002Fplaylist?list=PL2-dafEMk2A4ut2pyv0fSIXqOzXtBGkLj)\n- [Machine Learning for Developers For absolute beginners and fifth graders](https:\u002F\u002Fxyclade.github.io\u002FMachineLearning\u002F)\n- [dive-into-machine-learning](https:\u002F\u002Fgithub.com\u002Fhangtwenty\u002Fdive-into-machine-learning) Dive into Machine Learning with Python Jupyter notebook and scikit-learn\n- [An introduction to machine learning with scikit-learn](http:\u002F\u002Fscikit-learn.org\u002Fstable\u002Ftutorial\u002Fbasic\u002Ftutorial.html)\n- [Building powerful image classification models using very little data](https:\u002F\u002Fblog.keras.io\u002Fbuilding-powerful-image-classification-models-using-very-little-data.html)\n- [How HBO’s Silicon Valley built “Not Hotdog” with mobile TensorFlow, Keras & React Nativ](https:\u002F\u002Fmedium.com\u002F@timanglade\u002Fhow-hbos-silicon-valley-built-not-hotdog-with-mobile-tensorflow-keras-react-native-ef03260747f3)\n- [Hello World - Machine Learning Recipes #1](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=cKxRvEZd3Mw)\n- [An Intuitive Explanation of Convolutional Neural Networks](https:\u002F\u002Fujjwalkarn.me\u002F2016\u002F08\u002F11\u002Fintuitive-explanation-convnets\u002F) :rocket:\n- [A Quick Introduction to Neural Networks](https:\u002F\u002Fujjwalkarn.me\u002F2016\u002F08\u002F09\u002Fquick-intro-neural-networks\u002F)\n- [Machine Learning Algorithm for Flappy Bird using Neural Network and Genetic Algorithm](http:\u002F\u002Fwww.askforgametask.com\u002Ftutorial\u002Fmachine-learning-algorithm-flappy-bird\u002F)\n- [Understanding How Machines Learn, Through Prototyping](https:\u002F\u002Fmedium.com\u002Fbigtomorrow\u002Funderstanding-how-machines-learn-through-prototyping-9bdaa3ce7baa#.mura2rwy2) :rocket:\n- [Training on the device](http:\u002F\u002Fmachinethink.net\u002Fblog\u002Ftraining-on-device\u002F)\n- [Machine Learning for iOS](https:\u002F\u002Fwww.invasivecode.com\u002Fweblog\u002Fmachine-learning-swift-ios\u002F)\n- [The “hello world” of neural networks](http:\u002F\u002Fmachinethink.net\u002Fblog\u002Fthe-hello-world-of-neural-networks\u002F)\n- [Convolutional Neural Networks in iOS 10 and macOS](https:\u002F\u002Fwww.invasivecode.com\u002Fweblog\u002Fconvolutional-neural-networks-ios-10-macos-sierra\u002F)\n- [LearningMachineLearning](https:\u002F\u002Fgithub.com\u002Fgraceavery\u002FLearningMachineLearning) Swift implementation of \"Data Science From Scratch\" and http:\u002F\u002Fkarpathy.github.io\u002Fneuralnets\u002F\n- [The “hello world” of neural networks](http:\u002F\u002Fmachinethink.net\u002Fblog\u002Fthe-hello-world-of-neural-networks\u002F) 👶\n- [EmojiIntelligence](https:\u002F\u002Fgithub.com\u002FLuubra\u002FEmojiIntelligence) Neural Network built in Apple Playground using Swift 👶\n- [Machine Learning: End-to-end Classification](https:\u002F\u002Fwww.raywenderlich.com\u002F5554-machine-learning-end-to-end-classification)\n- [Machine Learning Zero to Hero (Google I\u002FO'19)](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=VwVg9jCtqaU&t=315s) :rocket:\n\n\n### Convolution neural network\n\n- [CS231n Winter 2016 Lecture 7 Convolutional Neural Networks](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=AQirPKrAyDg)\n- [How to build your own Neural Network from scratch in Python](https:\u002F\u002Ftowardsdatascience.com\u002Fhow-to-build-your-own-neural-network-from-scratch-in-python-68998a08e4f6) :star:\n- [How Convolutional Neural Networks work](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=FmpDIaiMIeA) :star:\n\n## Misc\n\n### Blogs\n\n- [Fritz Heartbeat](https:\u002F\u002Fheartbeat.fritz.ai\u002F) :star:\n\n### Create ML\n\n- [Introduction to Create ML: How to Train Your Own Machine Learning Model in Xcode 10](https:\u002F\u002Fwww.appcoda.com\u002Fcreate-ml\u002F)\n\n### ML Kit\n\n- [Introducing ML Kit](https:\u002F\u002Fdevelopers.googleblog.com\u002F2018\u002F05\u002Fintroducing-ml-kit.html)\n\n### Vision\n\n- [Vision](https:\u002F\u002Fdeveloper.apple.com\u002Fdocumentation\u002Fvision) Apply high-performance image analysis and computer vision techniques to identify faces, detect features, and classify scenes in images and video.\n- [Blog-Getting-Started-with-Vision](https:\u002F\u002Fgithub.com\u002Fjeffreybergier\u002FBlog-Getting-Started-with-Vision)\n- [Swift World: What’s new in iOS 11 — Vision](https:\u002F\u002Fmedium.com\u002Fcompileswift\u002Fswift-world-whats-new-in-ios-11-vision-456ba4156bad)\n\n### Natural Language Processing\n\n- [NSLinguisticTagger](https:\u002F\u002Fdeveloper.apple.com\u002Fdocumentation\u002Ffoundation\u002Fnslinguistictagger) Analyze natural language to tag part of speech and lexical class, identify proper names, perform lemmatization, and determine the language and script (orthography) of text.\n- [Linguistic Tagging](https:\u002F\u002Fwww.objc.io\u002Fissues\u002F7-foundation\u002Flinguistic-tagging\u002F)\n- [NSLinguisticTagger on NSHipster](http:\u002F\u002Fnshipster.com\u002Fnslinguistictagger\u002F)\n- [CoreLinguistics](https:\u002F\u002Fgithub.com\u002Frxwei\u002FCoreLinguistics) This repository contains some fundamental data structures for NLP.\n- [SwiftVerbalExpressions](https:\u002F\u002Fgithub.com\u002FVerbalExpressions\u002FSwiftVerbalExpressions) Swift Port of VerbalExpressions\n\n### Metal\n\n- [Metal](https:\u002F\u002Fdeveloper.apple.com\u002Fmetal\u002F)\n- [MPSCNNHelloWorld: Simple Digit Detection Convolution Neural Networks (CNN)](https:\u002F\u002Fdeveloper.apple.com\u002Flibrary\u002Fcontent\u002Fsamplecode\u002FMPSCNNHelloWorld\u002FIntroduction\u002FIntro.html)\n- [MetalImageRecognition: Performing Image Recognition](https:\u002F\u002Fdeveloper.apple.com\u002Flibrary\u002Fcontent\u002Fsamplecode\u002FMetalImageRecognition\u002FIntroduction\u002FIntro.html)\n- [Apple’s deep learning frameworks: BNNS vs. Metal CNN](http:\u002F\u002Fmachinethink.net\u002Fblog\u002Fapple-deep-learning-bnns-versus-metal-cnn\u002F)\n- [Forge](https:\u002F\u002Fgithub.com\u002Fhollance\u002FForge) A neural network toolkit for Metal\n\n### GamePlayKit\n\n- [GamePlayKit](https:\u002F\u002Fdeveloper.apple.com\u002Fdocumentation\u002Fgameplaykit)\n- [Project 34: Four in a Row](https:\u002F\u002Fwww.hackingwithswift.com\u002Fread\u002F34\u002Foverview)\n- [GKMinmaxStrategist: What does it take to build a TicTacToe AI?](http:\u002F\u002Ftilemapkit.com\u002F2015\u002F07\u002Fgkminmaxstrategist-build-tictactoe-ai\u002F)\n- [GameplayKit Tutorial: Artificial Intelligence](https:\u002F\u002Fwww.raywenderlich.com\u002F146407\u002Fgameplaykit-tutorial-artificial-intelligence)\n- [Gems of GameplayKit](https:\u002F\u002Fvimeo.com\u002Falbum\u002F4786409\u002Fvideo\u002F235143936)\n- [GameplayKit: Beyond Games](https:\u002F\u002Facademy.realm.io\u002Fposts\u002Fsash-zats-gameplaykit-beyond-games\u002F)\n\n### Courses\n\n- [6.S191: Introduction to Deep Learning](http:\u002F\u002Fintrotodeeplearning.com\u002Findex.html)\n- [Machine Learning](http:\u002F\u002Fintrotodeeplearning.com\u002Findex.html)\n- [Introduction - Intro to Machine Learning on Udacity](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=ICKBWIkfeJ8&list=PLAwxTw4SYaPkQXg8TkVdIvYv4HfLG7SiH)\n- [Machine Learning & Deep Learning Fundamentals by deeplizard](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=hfK_dvC-avg&list=PLZbbT5o_s2xq7LwI2y8_QtvuXZedL6tQU) :star: :star:\n\n### Interview\n\n- [41 Essential Machine Learning Interview Questions](https:\u002F\u002Fwww.springboard.com\u002Fblog\u002Fmachine-learning-interview-questions\u002F)\n\n### Other ML frameworks\n\n- [TensorSwift](https:\u002F\u002Fgithub.com\u002Fqoncept\u002FTensorSwift) A lightweight library to calculate tensors in Swift, which has similar APIs to TensorFlow's\n- [Swift-AI](https:\u002F\u002Fgithub.com\u002FSwift-AI\u002FSwift-AI) The Swift machine learning library.\n- [Swift-Brain](https:\u002F\u002Fgithub.com\u002Fvlall\u002FSwift-Brain) Artificial intelligence\u002Fmachine learning data structures and Swift algorithms for future iOS development. bayes theorem, neural networks, and more AI.\n- [Bender](https:\u002F\u002Fgithub.com\u002Fxmartlabs\u002FBender) Easily craft fast Neural Networks on iOS! Use TensorFlow models. Metal under the hood.\n- [BrainCore](https:\u002F\u002Fgithub.com\u002Faleph7\u002FBrainCore) The iOS and OS X neural network framework\n- [AIToolbox](https:\u002F\u002Fgithub.com\u002FKevinCoble\u002FAIToolbox) A toolbox of AI modules written in Swift: Graphs\u002FTrees, Support Vector Machines, Neural Networks, PCA, K-Means, Genetic Algorithms\n- [brain](https:\u002F\u002Fgithub.com\u002Fharthur\u002Fbrain) Neural networks in JavaScript\n- [TensorFlow](https:\u002F\u002Fwww.tensorflow.org\u002Fget_started\u002Fmnist\u002Fbeginners) An open-source software library for Machine Intelligence\n- [incubator-predictionio](https:\u002F\u002Fgithub.com\u002Fapache\u002Fincubator-predictionio) PredictionIO, a machine learning server for developers and ML engineers. Built on Apache Spark, HBase and Spray.\n- [Caffe](http:\u002F\u002Fcaffe.berkeleyvision.org\u002F) Deep learning framework by BAIR\n- [Torch](http:\u002F\u002Ftorch.ch\u002F) A SCIENTIFIC COMPUTING FRAMEWORK FOR LUAJIT\n- [Theano](http:\u002F\u002Fwww.deeplearning.net\u002Fsoftware\u002Ftheano\u002F) Theano is a Python library that allows you to define, optimize, and evaluate mathematical expressions involving multi-dimensional arrays efficiently\n- [CNTK](https:\u002F\u002Fgithub.com\u002FMicrosoft\u002FCNTK) Microsoft Cognitive Toolkit (CNTK), an open source deep-learning toolkit\n- [MXNet](http:\u002F\u002Fmxnet.io\u002F) Lightweight, Portable, Flexible Distributed\u002FMobile Deep Learning\n\n### Accelerate\n\n- [Accelerate-in-Swift](https:\u002F\u002Fgithub.com\u002Fhyperjeff\u002FAccelerate-in-Swift) Swift example codes for the Accelerate.framework\n- [Surge](https:\u002F\u002Fgithub.com\u002Fmattt\u002FSurge) A Swift library that uses the Accelerate framework to provide high-performance functions for matrix math, digital signal processing, and image manipulation.\n\n### Statistics\n\n- [SigmaSwiftStatistics](https:\u002F\u002Fgithub.com\u002Fevgenyneu\u002FSigmaSwiftStatistics) A collection of functions for statistical calculation written in Swift\n\n### Services\n\n- [Watson](https:\u002F\u002Fwww.ibm.com\u002Fwatson\u002Fdevelopercloud\u002F) Enable cognitive computing features in your app using IBM Watson's Language, Vision, Speech and Data APIs.\n- [wit.ai](https:\u002F\u002Fwit.ai\u002F) Natural Language for Developers\n- [Cloud Machine Learning Engine](https:\u002F\u002Fcloud.google.com\u002Fml-engine\u002F) Machine Learning on any data, any size\n- [Cloud Vision API](https:\u002F\u002Fcloud.google.com\u002Fvision\u002F) Derive insight from images with our powerful Cloud Vision API\n- [Amazon Machine Learning](https:\u002F\u002Faws.amazon.com\u002Fdocumentation\u002Fmachine-learning\u002F) Amazon Machine Learning makes it easy for developers to build smart applications, including applications for fraud detection, demand forecasting, targeted marketing, and click prediction\n- [api.ai](https:\u002F\u002Fapi.ai\u002F) Build brand-unique, natural language interactions for bots, applications, services, and devices.\n- [clarifai](https:\u002F\u002Fdeveloper.clarifai.com\u002F) Build amazing apps with the world’s best image and video recognition API.\n- [openml](https:\u002F\u002Fwww.openml.org\u002F) Exploring machine learning together\n- [Lobe](https:\u002F\u002Flobe.ai\u002F) Deep learning made simple\n- [Comparing Machine Learning (ML) Services from Various Cloud ML Service Providers](https:\u002F\u002Fmedium.com\u002F@tanyathakur6\u002Fcomparing-machine-learning-ml-services-from-various-cloud-ml-service-providers-63c8a2626cb6)\n\n### Text Recognition\n\n- [Tesseract OCR Tutorial](https:\u002F\u002Fwww.raywenderlich.com\u002F93276\u002Fimplementing-tesseract-ocr-ios)\n- [Tesseract-OCR-iOS](https:\u002F\u002Fgithub.com\u002Fgali8\u002FTesseract-OCR-iOS) Tesseract OCR iOS is a Framework for iOS7+, compiled also for armv7s and arm64.\n- [tesseract.js](https:\u002F\u002Fgithub.com\u002Fnaptha\u002Ftesseract.js) Pure Javascript OCR for 62 Languages\n\n### Speech Recognition\n\n- [Speech](https:\u002F\u002Fdeveloper.apple.com\u002Fdocumentation\u002Fspeech)\n- [Using the Speech Recognition API in iOS 10](https:\u002F\u002Fcode.tutsplus.com\u002Ftutorials\u002Fusing-the-speech-recognition-api-in-ios-10--cms-28032)\n- [Speech Recognition Tutorial for iOS](https:\u002F\u002Fwww.raywenderlich.com\u002F155752\u002Fspeech-recognition-tutorial-ios)\n- [CeedVocal](https:\u002F\u002Fgithub.com\u002Fcreaceed\u002FCeedVocal) Speech Recognition Library for iOS\n- [FluidAudio](https:\u002F\u002Fgithub.com\u002FFluidInference\u002FFluidAudio) Local audio AI SDK for Apple platforms with ASR, speaker diarization, VAD, and TTS optimized for Apple Neural Engine\n\n### Speech Synthesizer\n\n- [AVSpeechSynthesizer](https:\u002F\u002Fdeveloper.apple.com\u002Fdocumentation\u002Favfoundation\u002Favspeechsynthesizer) An object that produces synthesized speech from text utterances and provides controls for monitoring or controlling ongoing speech.\n\n### Artificial Intelligence\n\n- [The classic ELIZA chat bot in Swift.](https:\u002F\u002Fgist.github.com\u002Fhollance\u002Fbe70d0d7952066cb3160d36f33e5636f)\n- [Introduction to AI Programming for Games](https:\u002F\u002Fwww.raywenderlich.com\u002F24824\u002Fintroduction-to-ai-programming-for-games)\n\n### Google Cloud Platform for Machine Learning\n\n- [Machine Learning](https:\u002F\u002Fcloud.google.com\u002Fproducts\u002Fmachine-learning\u002F)\n- [Machine Learning APIs by Example (Google I\u002FO '17)](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=ETeeSYMGZn0)\n- [Adding Computer Vision to your iOS App](https:\u002F\u002Fmedium.com\u002F@srobtweets\u002Fadding-computer-vision-to-your-ios-app-66d6f540cdd2)\n\n### Others\n\n- [NotHotdog-Classifier](https:\u002F\u002Fgithub.com\u002Fkmather73\u002FNotHotdog-Classifier) What would you say if I told you there is a app on the market that tell you if you have a hotdog or not a hotdog.\n- [How HBO’s Silicon Valley built “Not Hotdog” with mobile TensorFlow, Keras & React Native](https:\u002F\u002Fhackernoon.com\u002Fhow-hbos-silicon-valley-built-not-hotdog-with-mobile-tensorflow-keras-react-native-ef03260747f3)\n","# 令人惊叹的机器学习 [![Awesome](https:\u002F\u002Fcdn.rawgit.com\u002Fsindresorhus\u002Fawesome\u002Fd7305f38d29fed78fa85652e3a63e154dd8e8829\u002Fmedia\u002Fbadge.svg)](https:\u002F\u002Fgithub.com\u002Fsindresorhus\u002Fawesome)\n\n❤️ 支持我的应用 ❤️ \n\n- [Push Hero - 纯 Swift 原生 macOS 应用，用于测试推送通知](https:\u002F\u002Fonmyway133.com\u002Fpushhero)\n- [PastePal - 剪贴板、笔记和快捷方式管理器](https:\u002F\u002Fonmyway133.com\u002Fpastepal)\n- [Quick Check - 智能待办事项管理器](https:\u002F\u002Fonmyway133.com\u002Fquickcheck)\n- [Alias - 应用和文件快捷方式管理器](https:\u002F\u002Fonmyway133.com\u002Falias)\n- [我的其他应用](https:\u002F\u002Fonmyway133.com\u002Fapps\u002F)\n\n❤️❤️😇😍🤘❤️❤️\n\n我喜欢探索机器学习，但不想为了理解某些框架或 TensorFlow 而深入到 Python 或 JavaScript 等其他平台。幸运的是，在 WWDC 2017 上，苹果推出了 Core ML、Vision 和 ARKit，这让使用机器学习变得容易得多。借助所有预训练模型，我们可以构建出很棒的应用。先感受成果，再尝试探索更高级的主题和底层机制，这感觉真好 🤖\n\n本项目主要整理与 Core ML 和 Swift 相关的内容。如果你需要参考其他平台的相关资料，这里也有涉及。\n\n## 目录\n\n- [Core ML](#core-ml)\n- [TensorFlow](#tensorflow)\n- [Keras](#keras)\n- [Turi Create](#turi-create)\n- [机器学习](#machine-learning)\n- [杂项](#misc)\n\n## Core ML\n\n### 模型 :rocket:\n\n- [Awesome-CoreML-Models](https:\u002F\u002Fgithub.com\u002Flikedan\u002FAwesome-CoreML-Models) 最大的 Core ML 模型列表（适用于 iOS 11 及以上版本）\n- [caffe](https:\u002F\u002Fgithub.com\u002FBVLC\u002Fcaffe) Caffe：一个快速的开源深度学习框架。http:\u002F\u002Fcaffe.berkeleyvision.org\u002F\n- [deep-learning-models](https:\u002F\u002Fgithub.com\u002Ffchollet\u002Fdeep-learning-models) Keras 中用于流行深度学习模型的代码和权重文件。\n- [tensorflow models](https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fmodels) 使用 TensorFlow 构建的模型\n- [libSVM](https:\u002F\u002Fwww.csie.ntu.edu.tw\u002F~cjlin\u002Flibsvm\u002F) 支持向量机库\n- [scikit-learn](http:\u002F\u002Fscikit-learn.org\u002F) Python 中的机器学习工具\n- [xgboost](https:\u002F\u002Fgithub.com\u002Fdmlc\u002Fxgboost) 一个可扩展、可移植且分布式的梯度提升库（GBDT、GBRT 或 GBM），支持 Python、R、Java、Scala、C++ 等语言。可在单机、Hadoop、Spark、Flink 和 DataFlow 等环境中运行。\n- [Keras-Classification-Models](https:\u002F\u002Fgithub.com\u002Ftitu1994\u002FKeras-Classification-Models) 用于分类任务的 Keras 模型集合\n- [MobileNet-Caffe](https:\u002F\u002Fgithub.com\u002Fshicai\u002FMobileNet-Caffe) Google MobileNets 的 Caffe 实现\n- [ModelZoo](https:\u002F\u002Fgithub.com\u002Fcocoa-ai\u002FModelZoo) 一个用于共享 Core ML 模型的中央 GitHub 仓库\n- [StyleArt](https:\u002F\u002Fgithub.com\u002Fileafsolutions\u002FStyleArt) Style Art 库利用一组预训练的机器学习模型处理图像，并将其转换为艺术风格\n- [models](https:\u002F\u002Fgithub.com\u002FSarasra\u002Fmodels) 使用 TensorFlow 构建的模型和示例\n- [Core ML Store](https:\u002F\u002Fcoreml.store\u002F)\n\n### 工具\n\n- [coremltools](https:\u002F\u002Fpypi.python.org\u002Fpypi\u002Fcoremltools) 一个用于创建、检查和测试 .mlmodel 格式模型的 Python 包\n- [torch2coreml](https:\u002F\u002Fgithub.com\u002Fprisma-ai\u002Ftorch2coreml) 此工具可帮助将 Torch7 模型转换为 Apple CoreML 格式\n- [turicreate](https:\u002F\u002Fgithub.com\u002Fapple\u002Fturicreate) Turi Create 简化了自定义机器学习模型的开发过程。\n- [Netron](https:\u002F\u002Fgithub.com\u002Flutzroeder\u002FNetron) 神经网络和机器学习模型查看器\n- [onnx-coreml](https:\u002F\u002Fgithub.com\u002Fonnx\u002Fonnx-coreml) ONNX 到 CoreML 转换器\n- [tf-coreml](https:\u002F\u002Fgithub.com\u002Ftf-coreml\u002Ftf-coreml) TensorFlow 到 CoreML 转换器\n- [tensorwatch](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Ftensorwatch) 用于机器学习和数据科学的调试与可视化工具\n\n### 文章\n\n- [Swift 教程：iOS 11 中的原生机器学习与机器视觉](https:\u002F\u002Fhackernoon.com\u002Fswift-tutorial-native-machine-learning-and-machine-vision-in-ios-11-11e1e88aa397)\n- [如何为 CoreML 训练自己的模型](http:\u002F\u002Freza.codes\u002F2017-07-29\u002Fhow-to-train-your-own-dataset-for-coreml\u002F)\n- [使用 CoreML 和 TensorFlow 在 iOS 11 中实现智能手势识别](https:\u002F\u002Fhackernoon.com\u002Fsmart-gesture-recognition-in-ios-11-with-core-ml-and-tensorflow-1a0a92c99c51)\n- [用 CoreML 打造 DIY Prisma 应用](https:\u002F\u002Fblog.prismalabs.ai\u002Fdiy-prisma-app-with-coreml-6b4994cc99e1) :rocket:\n- [使用 scikit-learn 和 CoreML 构建音乐推荐引擎](https:\u002F\u002Fwww.agnosticdev.com\u002Fblog-entry\u002Fpython\u002Fusing-scikit-learn-and-coreml-create-music-recommendation-engine)\n- [用 Turi Create 和 Core ML 一 afternoon 搭建“不是热狗”应用](https:\u002F\u002Fheartbeat.fritz.ai\u002Fbuilding-not-hotdog-with-turi-create-and-core-ml-in-an-afternoon-a87fd1967d10)\n- [利用 TensorFlow 对象检测 API、ML Engine 和 Swift 构建泰勒·斯威夫特识别器](https:\u002F\u002Ftowardsdatascience.com\u002Fbuild-a-taylor-swift-detector-with-the-tensorflow-object-detection-api-ml-engine-and-swift-82707f5b4a56)\n- [在 iOS 中利用机器学习提升无障碍体验](https:\u002F\u002Fmedium.com\u002Fbuffer-engineering\u002Fleveraging-machine-learning-in-ios-for-improved-accessibility-fc7796c5326f)\n- [IBM Watson 服务与 CoreML 教程](https:\u002F\u002Fwww.raywenderlich.com\u002F190191\u002Fibm-watson-services-for-core-ml-tutorial)\n- [从 Keras 和 CoreML 开始机器学习](https:\u002F\u002Fwww.raywenderlich.com\u002F181760\u002Fbeginning-machine-learning-keras-core-ml)\n- [使用 CoreML 和 IBM Watson 服务检测威士忌品牌](https:\u002F\u002Fmartinmitrevski.com\u002F2018\u002F04\u002F14\u002Fdetecting-whisky-brands-with-core-ml-and-ibm-watson-services\u002F)\n- [在你的 iOS 应用中使用 IBM Watson 和 CoreML 检测复仇者联盟超级英雄](https:\u002F\u002Fmedium.com\u002Fflawless-app-stories\u002Fdetecting-avengers-superheroes-in-your-ios-app-with-ibm-watson-and-coreml-fe38e493a4d1)\n- [机器学习](https:\u002F\u002Fdeveloper.apple.com\u002Fmachine-learning\u002F) 使用机器学习打造更智能的应用。\n- [苹果机器学习期刊](https:\u002F\u002Fmachinelearning.apple.com\u002F)\n- [CoreML 介绍](https:\u002F\u002Fdeveloper.apple.com\u002Fvideos\u002Fplay\u002Fwwdc2017\u002F703\u002F)\n- [深入解析 CoreML](https:\u002F\u002Fdeveloper.apple.com\u002Fvideos\u002Fplay\u002Fwwdc2017\u002F710\u002F)\n- [CoreML 和 Vision：iOS 11 中的机器学习教程](https:\u002F\u002Fwww.raywenderlich.com\u002F164213\u002Fcoreml-and-vision-machine-learning-in-ios-11-tutorial)\n- [iOS 11：让每个人都能享受机器学习](http:\u002F\u002Fmachinethink.net\u002Fblog\u002Fios-11-machine-learning-for-everyone\u002F)\n- [Swift 开发者想知道的关于机器学习的一切](https:\u002F\u002Fnews.realm.io\u002Fnews\u002Fswift-developer-on-machine-learning-try-swift-2017-gallagher\u002F)\n- [将机器学习引入你的 iOS 应用](https:\u002F\u002Fnews.realm.io\u002Fnews\u002Faltconf-2017-meghan-kane-bringing-machine-learning-to-your-ios-apps\u002F)\n- [iOS 机器学习 API 的优缺点](http:\u002F\u002Fmachinethink.net\u002Fblog\u002Fmachine-learning-apis\u002F) :rocket:\n- [CoreML：面向 iOS 的机器学习](https:\u002F\u002Fwww.udacity.com\u002Fcourse\u002Fcore-ml--ud1038) :rocket:\n- [机器学习训练流程的快速启动](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=ugiPfm8ICZo)\n- [使用 Create ML 在 iOS 应用中检测肺炎](https:\u002F\u002Fheartbeat.fritz.ai\u002Fdetecting-pneumonia-in-an-ios-app-with-create-ml-5cff2a60a3d)\n- [如何在 Keras 中微调 ResNet，并通过 CoreML 在 iOS 应用中使用](https:\u002F\u002Fheartbeat.fritz.ai\u002Fhow-to-fine-tune-resnet-in-keras-and-use-it-in-an-ios-app-via-core-ml-ee7fd84c1b26)\n- [机器学习中常见的五大数据质量问题及快速检测方法](https:\u002F\u002Fdoordash.engineering\u002F2022\u002F09\u002F27\u002Ffive-common-data-quality-gotchas-in-machine-learning-and-how-to-detect-them-quickly\u002F)\n\n### 仓库\n\n- [Core-ML-Sample](https:\u002F\u002Fgithub.com\u002Fyulingtianxia\u002FCore-ML-Sample) 使用 CoreML 框架的演示\n- [UnsplashExplorer-CoreML](https:\u002F\u002Fgithub.com\u002Fahmetws\u002FUnsplashExplorer-CoreML) 结合 Unsplash API 的 CoreML 演示应用\n- [MNIST_DRAW](https:\u002F\u002Fgithub.com\u002Fhwchong\u002FMNIST_DRAW) 这是一个示例项目，展示了如何使用 Keras（TensorFlow）训练 MNIST 手写数字识别模型，并在 iOS 11 上通过 CoreML 进行推理。\n- [CocoaAI](https:\u002F\u002Fgithub.com\u002Fcocoa-ai\u002FCocoaAI) Cocoa 人工智能实验室 :rocket:\n- [complex-gestures-demo](https:\u002F\u002Fgithub.com\u002Fmitochrome\u002Fcomplex-gestures-demo) 展示如何在 iOS 应用中使用机器学习识别 13 种复杂手势\n- [Core-ML-Car-Recognition](https:\u002F\u002Fgithub.com\u002Flikedan\u002FCore-ML-Car-Recognition) 面向 CoreML 的汽车识别框架\n- [CoreML-in-ARKit](https:\u002F\u002Fgithub.com\u002Fhanleyweng\u002FCoreML-in-ARKit) 简单项目，用于在 AR 中检测物体并在其上方显示 3D 标签。这是一个使用 CoreML 的 ARKit 项目的基础模板\n- [trainer-mac](https:\u002F\u002Fgithub.com\u002Fmortenjust\u002Ftrainer-mac) 训练一个模型，然后生成一个完整的 Xcode 项目来使用该模型——无需编写任何代码\n- [GestureAI-CoreML-iOS](https:\u002F\u002Fgithub.com\u002Fakimach\u002FGestureAI-CoreML-iOS) 使用 CoreML 在 iOS 应用中实现手势识别\n- [visual-recognition-coreml](https:\u002F\u002Fgithub.com\u002Fwatson-developer-cloud\u002Fvisual-recognition-coreml) 使用 Watson Visual Recognition 和 CoreML 离线分类图像\n\n## TensorFlow\n\n### 文章\n\n- [开源 TensorFlow 模型（Google I\u002FO '17）](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=9ziVGkt8Gg4)\n- [Swift for TensorFlow](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=Yze693W4MaU)\n- [开始使用 TensorFlow 高级 API（Google I\u002FO '18）](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=tjsHSIG8I08&list=PLOU2XLYxmsIInFRc3M44HUTQc3b_YJ4-Y&index=37)\n- [在 iOS 上开始使用 TensorFlow](http:\u002F\u002Fmachinethink.net\u002Fblog\u002Ftensorflow-on-ios\u002F)\n- [介绍 TensorFlow.js：JavaScript 中的机器学习](https:\u002F\u002Fmedium.com\u002Ftensorflow\u002Fintroducing-tensorflow-js-machine-learning-in-javascript-bf3eab376db)\n- [TensorFlow for JavaScript（Google I\u002FO '18）](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=OmofOvMApTU)\n- [Colab](https:\u002F\u002Fcolab.research.google.com) Colaboratory 是谷歌的一个研究项目，旨在促进机器学习教育和研究的传播\n- [使用 TensorFlow 和 BNNS 将机器学习添加到你的 Mac 或 iOS 应用中](https:\u002F\u002Fwww.bignerdranch.com\u002Fblog\u002Fuse-tensorflow-and-bnns-to-add-machine-learning-to-your-mac-or-ios-app\u002F)\n\n### 仓库\n\n- [workshops](https:\u002F\u002Fgithub.com\u002Ftensorflow\u002Fworkshops) 几个练习，可用于 Google IO 2018 的活动中\n\n### 课程\n\n- [使用 TensorFlow API 的机器学习速成课程](https:\u002F\u002Fdevelopers.google.com\u002Fmachine-learning\u002Fcrash-course\u002F)\n- [弗朗索瓦·肖莱的《用 Python 学习深度学习》](https:\u002F\u002Fwww.amazon.com\u002FDeep-Learning-Python-Francois-Chollet\u002Fdp\u002F1617294438)\n- [向谷歌的机器学习专家学习](https:\u002F\u002Fai.google\u002Feducation)\n\n## Keras\n\n### 文章\n\n- [在 iOS 上使用 CoreML 运行 Keras 模型](https:\u002F\u002Fwww.pyimagesearch.com\u002F2018\u002F04\u002F23\u002Frunning-keras-models-on-ios-with-coreml\u002F)\n- [Keras 与卷积神经网络 (CNN)](https:\u002F\u002Fwww.pyimagesearch.com\u002F2018\u002F04\u002F16\u002Fkeras-and-convolutional-neural-networks-cnns\u002F)\n- [Keras 教程：使用预训练模型进行迁移学习]()https:\u002F\u002Fwww.learnopencv.com\u002Fkeras-tutorial-transfer-learning-using-pre-trained-models\u002F\n\n## Turi Create\n\n### 文章\n\n- [用 Turi Create 和 Core ML 在一个下午构建“不是热狗”应用](https:\u002F\u002Fhackernoon.com\u002Fbuilding-not-hotdog-with-turi-create-and-core-ml-in-an-afternoon-231b14738edf)\n- [使用 Turi Create 在 iOS 上进行自然语言处理](https:\u002F\u002Fwww.raywenderlich.com\u002F185515\u002Fnatural-language-processing-on-ios-with-turi-create)\n- [iOS 中的机器学习：Turi Create 和 CoreML](https:\u002F\u002Fmedium.com\u002Fflawless-app-stories\u002Fmachine-learning-in-ios-turi-create-and-coreml-5ddce0dc8e26)\n- [Turi Create 使用指南](https:\u002F\u002Fdeveloper.apple.com\u002Fvideos\u002Fplay\u002Fwwdc2018\u002F712\u002F)\n\n## 机器学习\n\n### 入门\n\n- [致 Towards Data Science 的感谢信](https:\u002F\u002Ftowardsdatascience.com\u002Fa-thank-you-note-to-towards-data-science-58b714a824f8)\n- [这就是为什么任何人都可以学习机器学习](https:\u002F\u002Fmedium.freecodecamp.org\u002Fthis-is-why-anyone-can-learn-machine-learning-a5333ee64dff)\n\n### 文章\n\n- [机器学习的可视化入门](http:\u002F\u002Fwww.r2d3.us\u002Fvisual-intro-to-machine-learning-part-1\u002F)\n- [机器学习真有趣！](https:\u002F\u002Fmedium.com\u002F@ageitgey\u002Fmachine-learning-is-fun-80ea3ec3c471)\n- [用简单英语解释的 10 个机器学习术语](http:\u002F\u002Fblog.aylien.com\u002F10-machine-learning-terms-explained-in-simple\u002F)\n- [一年学会机器学习](https:\u002F\u002Fmedium.com\u002Flearning-new-stuff\u002Fmachine-learning-in-a-year-cdb0b0ebd29c)\n- [机器学习自学资源](https:\u002F\u002Fragle.sanukcode.net\u002Farticles\u002Fmachine-learning-self-study-resources\u002F)\n- [如何学习机器学习](https:\u002F\u002Felitedatascience.com\u002Flearn-machine-learning)\n- [机器学习入门](https:\u002F\u002Fmedium.com\u002F@suffiyanz\u002Fgetting-started-with-machine-learning-f15df1c283ea)\n- [面向非技术人员的机器学习与人工智能指南](https:\u002F\u002Fmachinelearnings.co\u002Fa-humans-guide-to-machine-learning-e179f43b67a0)\n- [机器学习深度指南：概述、目标、学习类型和算法](http:\u002F\u002Fwww.innoarchitech.com\u002Fmachine-learning-an-in-depth-non-technical-guide\u002F)\n- [机器学习算法概览](http:\u002F\u002Fmachinelearningmastery.com\u002Fa-tour-of-machine-learning-algorithms\u002F)\n- [黑客的机器学习](https:\u002F\u002Fwww.youtube.com\u002Fplaylist?list=PL2-dafEMk2A4ut2pyv0fSIXqOzXtBGkLj)\n- [面向开发人员的机器学习：适合绝对新手和五年级学生](https:\u002F\u002Fxyclade.github.io\u002FMachineLearning\u002F)\n- [深入机器学习](https:\u002F\u002Fgithub.com\u002Fhangtwenty\u002Fdive-into-machine-learning)：使用 Python Jupyter Notebook 和 scikit-learn 深入学习机器学习\n- [使用 scikit-learn 学习机器学习的入门教程](http:\u002F\u002Fscikit-learn.org\u002Fstable\u002Ftutorial\u002Fbasic\u002Ftutorial.html)\n- [使用极少数据构建强大的图像分类模型](https:\u002F\u002Fblog.keras.io\u002Fbuilding-powerful-image-classification-models-using-very-little-data.html)\n- [HBO《硅谷》如何用移动 TensorFlow、Keras 和 React Native 构建“不是热狗”](https:\u002F\u002Fmedium.com\u002F@timanglade\u002Fhow-hbos-silicon-valley-built-not-hotdog-with-mobile-tensorflow-keras-react-native-ef03260747f3)\n- [Hello World - 机器学习食谱 #1](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=cKxRvEZd3Mw)\n- [卷积神经网络的直观解释](https:\u002F\u002Fujjwalkarn.me\u002F2016\u002F08\u002F11\u002Fintuitive-explanation-convnets\u002F) :rocket:\n- [神经网络快速入门](https:\u002F\u002Fujjwalkarn.me\u002F2016\u002F08\u002F09\u002Fquick-intro-neural-networks\u002F)\n- [使用神经网络和遗传算法实现 Flappy Bird 的机器学习算法](http:\u002F\u002Fwww.askforgametask.com\u002Ftutorial\u002Fmachine-learning-algorithm-flappy-bird\u002F)\n- [通过原型设计理解机器如何学习](https:\u002F\u002Fmedium.com\u002Fbigtomorrow\u002Funderstanding-how-machines-learn-through-prototyping-9bdaa3ce7baa#.mura2rwy2) :rocket:\n- [设备端训练](http:\u002F\u002Fmachinethink.net\u002Fblog\u002Ftraining-on-device\u002F)\n- [iOS 中的机器学习](https:\u002F\u002Fwww.invasivecode.com\u002Fweblog\u002Fmachine-learning-swift-ios\u002F)\n- [神经网络的“Hello World”](http:\u002F\u002Fmachinethink.net\u002Fblog\u002Fthe-hello-world-of-neural-networks\u002F) 👶\n- [EmojiIntelligence](https:\u002F\u002Fgithub.com\u002FLuubra\u002FEmojiIntelligence)：使用 Swift 在 Apple Playground 中构建的神经网络 👶\n- [机器学习：端到端分类](https:\u002F\u002Fwww.raywenderlich.com\u002F5554-machine-learning-end-to-end-classification)\n- [机器学习从零到英雄（Google I\u002FO'19）](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=VwVg9jCtqaU&t=315s) :rocket:\n\n\n### 卷积神经网络\n\n- [CS231n 2016 冬季第 7 讲：卷积神经网络](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=AQirPKrAyDg)\n- [如何用 Python 从头开始构建自己的神经网络](https:\u002F\u002Ftowardsdatascience.com\u002Fhow-to-build-your-own-neural-network-from-scratch-in-python-68998a08e4f6) :star:\n- [卷积神经网络的工作原理](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=FmpDIaiMIeA) :star:\n\n## 杂项\n\n### 博客\n\n- [Fritz Heartbeat](https:\u002F\u002Fheartbeat.fritz.ai\u002F) :star:\n\n### Create ML\n\n- [Create ML 入门：如何在 Xcode 10 中训练你自己的机器学习模型](https:\u002F\u002Fwww.appcoda.com\u002Fcreate-ml\u002F)\n\n### ML Kit\n\n- [推出 ML Kit](https:\u002F\u002Fdevelopers.googleblog.com\u002F2018\u002F05\u002Fintroducing-ml-kit.html)\n\n### Vision\n\n- [Vision](https:\u002F\u002Fdeveloper.apple.com\u002Fdocumentation\u002Fvision)：应用高性能图像分析和计算机视觉技术，以识别面部、检测特征并分类图像和视频中的场景。\n- [Vision 入门博客](https:\u002F\u002Fgithub.com\u002Fjeffreybergier\u002FBlog-Getting-Started-with-Vision)\n- [Swift World：iOS 11 新功能——Vision](https:\u002F\u002Fmedium.com\u002Fcompileswift\u002Fswift-world-whats-new-in-ios-11-vision-456ba4156bad)\n\n### 自然語言處理\n\n- [NSLinguisticTagger](https:\u002F\u002Fdeveloper.apple.com\u002Fdocumentation\u002Ffoundation\u002Fnslinguistictagger) 分析自然語言，標記詞性與詞類、識別專有名詞、進行詞形還原，並判斷文本的語言及書寫系統（正字法）。\n- [語言標記](https:\u002F\u002Fwww.objc.io\u002Fissues\u002F7-foundation\u002Flinguistic-tagging\u002F)\n- [NSLinguisticTagger 在 NSHipster 上](http:\u002F\u002Fnshipster.com\u002Fnslinguistictagger\u002F)\n- [CoreLinguistics](https:\u002F\u002Fgithub.com\u002Frxwei\u002FCoreLinguistics) 此倉庫包含一些用於自然語言處理的基本數據結構。\n- [SwiftVerbalExpressions](https:\u002F\u002Fgithub.com\u002FVerbalExpressions\u002FSwiftVerbalExpressions) VerbalExpressions 的 Swift 版本\n\n### Metal\n\n- [Metal](https:\u002F\u002Fdeveloper.apple.com\u002Fmetal\u002F)\n- [MPSCNNHelloWorld：簡單的數字檢測卷積神經網絡 (CNN)](https:\u002F\u002Fdeveloper.apple.com\u002Flibrary\u002Fcontent\u002Fsamplecode\u002FMPSCNNHelloWorld\u002FIntroduction\u002FIntro.html)\n- [MetalImageRecognition：執行圖像識別](https:\u002F\u002Fdeveloper.apple.com\u002Flibrary\u002Fcontent\u002Fsamplecode\u002FMetalImageRecognition\u002FIntroduction\u002FIntro.html)\n- [Apple 的深度學習框架：BNNS 對比 Metal CNN](http:\u002F\u002Fmachinethink.net\u002Fblog\u002Fapple-deep-learning-bnns-versus-metal-cnn\u002F)\n- [Forge](https:\u002F\u002Fgithub.com\u002Fhollance\u002FForge) 一個用於 Metal 的神經網絡工具包\n\n### GamePlayKit\n\n- [GamePlayKit](https:\u002F\u002Fdeveloper.apple.com\u002Fdocumentation\u002Fgameplaykit)\n- [專案 34：四子連線](https:\u002F\u002Fwww.hackingwithswift.com\u002Fread\u002F34\u002Foverview)\n- [GKMinmaxStrategist：打造井字遊戲 AI 需要什麼？](http:\u002F\u002Ftilemapkit.com\u002F2015\u002F07\u002Fgkminmaxstrategist-build-tictactoe-ai\u002F)\n- [GameplayKit 教程：人工智能](https:\u002F\u002Fwww.raywenderlich.com\u002F146407\u002Fgameplaykit-tutorial-artificial-intelligence)\n- [GameplayKit 精華](https:\u002F\u002Fvimeo.com\u002Falbum\u002F4786409\u002Fvideo\u002F235143936)\n- [GameplayKit：超越遊戲](https:\u002F\u002Facademy.realm.io\u002Fposts\u002Fsash-zats-gameplaykit-beyond-games\u002F)\n\n### 課程\n\n- [6.S191：深度學習入門](http:\u002F\u002Fintrotodeeplearning.com\u002Findex.html)\n- [機器學習](http:\u002F\u002Fintrotodeeplearning.com\u002Findex.html)\n- [導論——Udacity 機器學習入門](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=ICKBWIkfeJ8&list=PLAwxTw4SYaPkQXg8TkVdIvYv4HfLG7SiH)\n- [deeplizard 的機器學習與深度學習基礎](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=hfK_dvC-avg&list=PLZbbT5o_s2xq7LwI2y8_QtvuXZedL6tQU) :star: :star:\n\n### 面試\n\n- [41 個必備的機器學習面試問題](https:\u002F\u002Fwww.springboard.com\u002Fblog\u002Fmachine-learning-interview-questions\u002F)\n\n### 其他 ML 框架\n\n- [TensorSwift](https:\u002F\u002Fgithub.com\u002Fqoncept\u002FTensorSwift) 一個輕量級的 Swift 庫，用於計算張量，其 API 與 TensorFlow 相似。\n- [Swift-AI](https:\u002F\u002Fgithub.com\u002FSwift-AI\u002FSwift-AI) Swift 機器學習庫。\n- [Swift-Brain](https:\u002F\u002Fgithub.com\u002Fvlall\u002FSwift-Brain) 用於未來 iOS 開發的人工智能\u002F機器學習數據結構及 Swift 算法。包括貝葉斯定理、神經網絡等 AI 技術。\n- [Bender](https:\u002F\u002Fgithub.com\u002Fxmartlabs\u002FBender) 在 iOS 上輕鬆構建快速神經網絡！使用 TensorFlow 模型，底層採用 Metal。\n- [BrainCore](https:\u002F\u002Fgithub.com\u002Faleph7\u002FBrainCore) iOS 和 OS X 平台上的神經網絡框架。\n- [AIToolbox](https:\u002F\u002Fgithub.com\u002FKevinCoble\u002FAIToolbox) 一個用 Swift 寫成的 AI 模塊工具箱：圖\u002F樹、支持向量機、神經網絡、主成分分析、K-均值聚類、遺傳算法等。\n- [brain](https:\u002F\u002Fgithub.com\u002Fharthur\u002Fbrain) JavaScript 中的神經網絡。\n- [TensorFlow](https:\u002F\u002Fwww.tensorflow.org\u002Fget_started\u002Fmnist\u002Fbeginners) 一個用於機器智慧的開源軟體庫。\n- [incubator-predictionio](https:\u002F\u002Fgithub.com\u002Fapache\u002Fincubator-predictionio) PredictionIO 是一款面向開發者和機器學習工程師的機器學習伺服器，基於 Apache Spark、HBase 和 Spray 構建。\n- [Caffe](http:\u002F\u002Fcaffe.berkeleyvision.org\u002F) BAIR 發布的深度學習框架。\n- [Torch](http:\u002F\u002Ftorch.ch\u002F) 一個用於 LuaJIT 的科學計算框架。\n- [Theano](http:\u002F\u002Fwww.deeplearning.net\u002Fsoftware\u002Ftheano\u002F) Theano 是一個 Python 庫，可高效地定義、優化和評估涉及多維陣列的數學表達式。\n- [CNTK](https:\u002F\u002Fgithub.com\u002FMicrosoft\u002FCNTK) Microsoft Cognitive Toolkit (CNTK)，一個開源深度學習工具包。\n- [MXNet](http:\u002F\u002Fmxnet.io\u002F) 輕量級、可移植、靈活的分布式\u002F移動端深度學習框架。\n\n### Accelerate\n\n- [Accelerate-in-Swift](https:\u002F\u002Fgithub.com\u002Fhyperjeff\u002FAccelerate-in-Swift) Accelerate.framework 的 Swift 示例代碼。\n- [Surge](https:\u002F\u002Fgithub.com\u002Fmattt\u002FSurge) 一個使用 Accelerate 框架的 Swift 庫，提供高效的矩陣運算、數位信號處理和圖像操作功能。\n\n### 統計學\n\n- [SigmaSwiftStatistics](https:\u002F\u002Fgithub.com\u002Fevgenyneu\u002FSigmaSwiftStatistics) 一個用 Swift 寫成的統計計算函數集合。\n\n### 服務\n\n- [Watson](https:\u002F\u002Fwww.ibm.com\u002Fwatson\u002Fdevelopercloud\u002F) 使用 IBM Watson 的語言、視覺、語音和數據 API，在您的應用中實現認知計算功能。\n- [wit.ai](https:\u002F\u002Fwit.ai\u002F) 面向開發者的自然語言處理平台。\n- [Cloud Machine Learning Engine](https:\u002F\u002Fcloud.google.com\u002Fml-engine\u002F) 可在任何規模的數據上進行機器學習。\n- [Cloud Vision API](https:\u002F\u002Fcloud.google.com\u002Fvision\u002F) 利用我們強大的 Cloud Vision API 從圖像中獲取洞見。\n- [Amazon Machine Learning](https:\u002F\u002Faws.amazon.com\u002Fdocumentation\u002Fmachine-learning\u002F) Amazon Machine Learning 讓開發者能夠輕鬆構建智能應用程序，包括欺詐檢測、需求預測、目標行銷和點擊預測等功能。\n- [api.ai](https:\u002F\u002Fapi.ai\u002F) 為機器人、應用程序、服務和設備構建獨特的品牌化自然語言交互。\n- [clarifai](https:\u002F\u002Fdeveloper.clarifai.com\u002F) 使用全球最佳的圖像和視頻識別 API 打造令人驚嘆的應用程序。\n- [openml](https:\u002F\u002Fwww.openml.org\u002F) 一起探索機器學習。\n- [Lobe](https:\u002F\u002Flobe.ai\u002F) 讓深度學習變得簡單。\n- [比較各雲端 ML 服務提供商的機器學習 (ML) 服務](https:\u002F\u002Fmedium.com\u002F@tanyathakur6\u002Fcomparing-machine-learning-ml-services-from-various-cloud-ml-service-providers-63c8a2626cb6)\n\n### 文字識別\n\n- [Tesseract OCR 教程](https:\u002F\u002Fwww.raywenderlich.com\u002F93276\u002Fimplementing-tesseract-ocr-ios)\n- [Tesseract-OCR-iOS](https:\u002F\u002Fgithub.com\u002Fgali8\u002FTesseract-OCR-iOS) Tesseract OCR iOS 是一個適用於 iOS7+ 的框架，同時針對 armv7s 和 arm64 架構編譯。\n- [tesseract.js](https:\u002F\u002Fgithub.com\u002Fnaptha\u002Ftesseract.js) 純 JavaScript OCR，支援 62 種語言。\n\n### 语音识别\n\n- [Speech](https:\u002F\u002Fdeveloper.apple.com\u002Fdocumentation\u002Fspeech)\n- [在 iOS 10 中使用语音识别 API](https:\u002F\u002Fcode.tutsplus.com\u002Ftutorials\u002Fusing-the-speech-recognition-api-in-ios-10--cms-28032)\n- [iOS 语音识别教程](https:\u002F\u002Fwww.raywenderlich.com\u002F155752\u002Fspeech-recognition-tutorial-ios)\n- [CeedVocal](https:\u002F\u002Fgithub.com\u002Fcreaceed\u002FCeedVocal) 适用于 iOS 的语音识别库\n- [FluidAudio](https:\u002F\u002Fgithub.com\u002FFluidInference\u002FFluidAudio) 面向 Apple 平台的本地音频 AI SDK，支持自动语音识别、说话人分离、语音活动检测和文本转语音功能，并针对 Apple 神经引擎进行了优化\n\n### 语音合成器\n\n- [AVSpeechSynthesizer](https:\u002F\u002Fdeveloper.apple.com\u002Fdocumentation\u002Favfoundation\u002Favspeechsynthesizer) 用于将文本转换为合成语音的对象，并提供用于监控或控制正在进行的语音输出的控件。\n\n### 人工智能\n\n- [用 Swift 实现的经典 ELIZA 聊天机器人](https:\u002F\u002Fgist.github.com\u002Fhollance\u002Fbe70d0d7952066cb3160d36f33e5636f)\n- [游戏人工智能编程入门](https:\u002F\u002Fwww.raywenderlich.com\u002F24824\u002Fintroduction-to-ai-programming-for-games)\n\n### Google Cloud Platform 机器学习服务\n\n- [机器学习](https:\u002F\u002Fcloud.google.com\u002Fproducts\u002Fmachine-learning\u002F)\n- [通过示例了解机器学习 API（Google I\u002FO 2017）](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=ETeeSYMGZn0)\n- [为你的 iOS 应用添加计算机视觉功能](https:\u002F\u002Fmedium.com\u002F@srobtweets\u002Fadding-computer-vision-to-your-ios-app-66d6f540cdd2)\n\n### 其他\n\n- [NotHotdog 分类器](https:\u002F\u002Fgithub.com\u002Fkmather73\u002FNotHotdog-Classifier) 如果我告诉你市面上有一款应用可以判断你拿的是热狗还是非热狗，你会作何感想？\n- [HBO《硅谷》如何利用移动 TensorFlow、Keras 和 React Native 构建“Not Hotdog”](https:\u002F\u002Fhackernoon.com\u002Fhow-hbos-silicon-valley-built-not-hotdog-with-mobile-tensorflow-keras-react-native-ef03260747f3)","# awesome-machine-learning 快速上手指南\n\n`awesome-machine-learning` 并非一个单一的可安装软件库，而是一个精选的机器学习资源列表（Awesome List），主要聚焦于 **Apple Core ML**、**Swift** 生态以及相关的模型转换工具。本指南将帮助你搭建开发环境，并演示如何利用列表中的核心工具将机器学习模型集成到 iOS\u002FmacOS 应用中。\n\n## 环境准备\n\n在开始之前，请确保你的开发环境满足以下要求：\n\n*   **操作系统**: macOS (推荐最新稳定版)\n*   **开发工具**: Xcode 14.0 或更高版本（包含 Swift 5.7+）\n*   **Python 环境**: Python 3.8 - 3.10（用于模型训练与转换）\n    *   建议使用 `conda` 或 `venv` 创建独立虚拟环境。\n*   **前置依赖库**:\n    *   `coremltools`: Apple 官方提供的模型转换与检查工具。\n    *   `tensorflow` 或 `keras` \u002F `pytorch`: 用于加载预训练模型。\n    *   `scikit-learn`: 用于传统机器学习算法。\n\n**国内加速建议**：\n在安装 Python 依赖时，推荐使用清华源或阿里源以提升下载速度：\n```bash\npip config set global.index-url https:\u002F\u002Fpypi.tuna.tsinghua.edu.cn\u002Fsimple\n```\n\n## 安装步骤\n\n### 1. 安装 Python 依赖\n打开终端，创建并激活虚拟环境，然后安装核心工具链：\n\n```bash\npython3 -m venv ml_env\nsource ml_env\u002Fbin\u002Factivate\n\n# 安装 coremltools 及常用深度学习框架\npip install coremltools tensorflow keras scikit-learn\n```\n\n### 2. 获取资源列表\n克隆该仓库以浏览推荐的模型、教程和开源项目代码：\n\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002Flikedan\u002FAwesome-CoreML-Models.git\n# 或者主列表\ngit clone https:\u002F\u002Fgithub.com\u002Fonmyway133\u002Fawesome-machine-learning.git\n```\n\n### 3. 下载预训练模型\n你可以直接从列表中推荐的仓库（如 [Awesome-CoreML-Models](https:\u002F\u002Fgithub.com\u002Flikedan\u002FAwesome-CoreML-Models)）下载 `.mlmodel` 文件，或者使用下方示例自行转换。\n\n## 基本使用\n\n以下示例演示如何使用 `coremltools` 将一个简单的 Keras 模型转换为 Apple Core ML 格式 (`.mlmodel`)，以便在 iOS 应用中使用。\n\n### 1. 构建并转换模型 (Python)\n\n创建一个名为 `convert_model.py` 的文件：\n\n```python\nimport coremltools as ct\nfrom tensorflow import keras\nfrom tensorflow.keras import layers\n\n# 1. 定义一个简单的 Keras 模型 (例如用于 MNIST 手写数字识别)\nmodel = keras.Sequential([\n    layers.Flatten(input_shape=(28, 28)),\n    layers.Dense(128, activation='relu'),\n    layers.Dense(10, activation='softmax')\n])\n\nmodel.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])\n\n# 注意：实际使用中需先 model.fit() 进行训练，此处仅为演示结构\n# model.fit(x_train, y_train, epochs=5) \n\n# 2. 转换为 Core ML 格式\n# input_name: 模型在 iOS 端接收数据的名称\n# class_labels: 如果是分类任务，指定类别标签\nmlmodel = ct.convert(model, source='tensorflow')\n\n# 3. 保存为 .mlmodel 文件\nmlmodel.save('MySimpleModel.mlmodel')\n\nprint(\"转换成功！文件已保存为 MySimpleModel.mlmodel\")\n```\n\n运行脚本：\n```bash\npython convert_model.py\n```\n\n### 2. 在 Xcode 中集成 (Swift)\n\n1.  将生成的 `MySimpleModel.mlmodel` 文件拖入 Xcode 项目的导航栏。\n2.  Xcode 会自动编译该模型并生成对应的 Swift 类（类名通常为文件名，如 `MySimpleModel`）。\n3.  在 Swift 代码中调用模型：\n\n```swift\nimport CoreML\nimport Vision\n\n\u002F\u002F 初始化模型\nguard let model = try? MySimpleModel(configuration: MLModelConfiguration()) else {\n    fatalError(\"无法加载模型\")\n}\n\n\u002F\u002F 准备输入数据 (假设输入为 28x28 的图像像素数据)\n\u002F\u002F 这里需要使用 MLMultiArray 处理数据，具体取决于模型输入类型\nlet input = MySimpleModelInput(imageData: myPixelData) \n\ndo {\n    \u002F\u002F 进行预测\n    let output = try model.prediction(input: input)\n    let label = output.classLabel\n    print(\"预测结果: \\(label)\")\n} catch {\n    print(\"预测失败: \\(error)\")\n}\n```\n\n### 3. 使用现有开源项目\n如果你想直接体验完整应用，可以克隆列表中的示例项目，例如手势识别或图像分类 Demo：\n\n```bash\n# 示例：克隆一个 Core ML 手势识别 Demo\ngit clone https:\u002F\u002Fgithub.com\u002Fakimach\u002FGestureAI-CoreML-iOS.git\ncd GestureAI-CoreML-iOS\nopen GestureAI.xcodeproj\n```\n直接在 Xcode 中运行即可在模拟器或真机上测试效果。","一位专注于 iOS 开发的独立开发者，希望在不深入 Python 或 TensorFlow 底层细节的前提下，快速为自家应用集成图像风格迁移功能。\n\n### 没有 awesome-machine-learning 时\n- **跨平台门槛高**：被迫花费数周学习 Python 环境和 TensorFlow 框架，偏离了 Swift 原生开发的主航道。\n- **模型寻找困难**：在海量开源项目中盲目搜索适配 Core ML 的预训练模型，难以验证其兼容性与效果。\n- **转换流程复杂**：缺乏明确的工具指引，不知道如何将现有的 Caffe 或 Keras 模型高效转换为 .mlmodel 格式。\n- **调试无从下手**：面对模型集成后的黑盒运行状态，缺少可视化工具来诊断神经网络结构或性能瓶颈。\n\n### 使用 awesome-machine-learning 后\n- **聚焦原生技术**：直接获取基于 Core ML 和 Swift 的精选资源清单，无需涉足其他编程语言即可上手。\n- **模型即取即用**：通过列表中推荐的 Awesome-CoreML-Models 和 ModelZoo，迅速找到并下载成熟的风格迁移模型。\n- **转换工具现成**：利用收录的 coremltools、tf-coreml 等转换利器，一键将主流框架模型转为苹果生态专用格式。\n- **可视化辅助**：借助 Netron 等工具直观查看模型层级结构，快速定位集成问题并优化应用表现。\n\nawesome-machine-learning 让 iOS 开发者能够跳过繁琐的跨平台学习曲线，直接站在巨人的肩膀上构建高质量的智能应用。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fonmyway133_awesome-machine-learning_dbf78054.png","onmyway133","Khoa","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Fonmyway133_0040a976.jpg","Building https:\u002F\u002Findiegoodies.com\u002F","Indie Goodies","Oslo, Norway","onmyway133@gmail.com","https:\u002F\u002Fonmyway133.com\u002F","https:\u002F\u002Fgithub.com\u002Fonmyway133",null,812,102,"2026-04-13T15:54:10","MIT",1,"macOS, iOS","未说明",{"notes":91,"python":92,"dependencies":93},"该项目主要是一个面向 Apple 生态系统（macOS\u002FiOS）的机器学习资源列表，核心开发语言为 Swift。虽然列出了 TensorFlow、Keras、scikit-learn 等 Python 库，但主要用于模型训练或转换为 Core ML 格式 (.mlmodel)，而非直接作为运行环境依赖。实际运行推理主要在 iOS 11+ 或 macOS 设备上利用 Apple Neural Engine 或 CPU\u002FGPU 进行，无需特定 NVIDIA GPU 或 CUDA 环境。","未说明 (核心工具链基于 Swift\u002FXcode，部分转换工具如 coremltools 需 Python)",[94,95,96,97,98,99,100,101,102],"Xcode","Core ML","Vision","ARKit","Swift","coremltools","Turi Create","TensorFlow","Keras",[15,14,35,13,16],[105,106,107,108,109,110,111,112,113,114],"core-ml","machine","learning","model","ai","vision","language","processing","augmented","reality","2026-03-27T02:49:30.150509","2026-04-14T12:30:10.802692",[],[]]