# OmniXRI TinyML 小學堂 (2025) 第 1 講 微型人工智慧簡介 <img src="https://raw.githubusercontent.com/OmniXRI/TinyML_Course_2025/refs/heads/main/images/2025_TinyML%E5%B0%8F%E5%AD%B8%E5%A0%82%E8%AA%B2%E7%A8%8B%E7%9B%B4%E6%92%AD%E5%9C%96%E7%A4%BA_Ch1.JPG" alt="" width="640"> <span style="color:#FF0000;">**本課程完全免費,請勿移作商業用途!更多課程內容請參考文末教學資源連結。歡迎追蹤、留言、訂閱、點讚、轉發,讓更多需要的朋友也能一起學習。**</span> **直播日期: 2025/03/12 20:00-21:20** **課程直播連結: https://youtu.be/3WGp3QFOLMw** **課程簡報名稱: [20250312_TInyML_Course_Ch01_OmniXRI_Jack.pdf](https://github.com/OmniXRI/TinyML_Course_2025/tree/main/Ch01_TinyML_Introduction)** ## 課程內容 1.1. 智慧物聯網與微型人工智慧 * 傳統物聯網 (IoT) 工作流程 * 工業物聯網 (IIoT) 架構 * 智慧物聯網 (AIoT) 架構 * 智慧健康(穿戴式)物聯網 * 邊緣智慧 vs. 生成智慧 * Edge AI 是什麼? * 邊緣智慧 (Edge AI) 誰說了算? * 何謂微型人工智慧 (TinyML) * TinyML 相關論文 * MCU等級TinyML常見應用及限制 * 智慧物聯網 vs. 微型人工智慧 1.2. TinyML 常見應用 * [TinyML案例分享(技術分類)](https://github.com/user-attachments/assets/9ae2f439-8c1d-40ed-999e-5e8aafdb4757) * [TinyML案例分享 ─ Hackster.io](https://www.hackster.io/search?q=tinyml&i=projects) * [TinyML案例分享 ─ Edge Impulse Blog](https://www.edgeimpulse.com/blog/) * [TinyML案例分享 ─ Arduino TinyML Blog](https://search.arduino.cc/search?tab=&q=tinyML) * [TinyML案例分享 ─ Edge AI基金會(Youtube)](https://www.youtube.com/@edgeaifoundation/videos) 1.3. TinyML 常見開發板 * MCU vs. MPU / SoC * MCU等級主要核心晶片分類 * 常見tinyML開發板(智能感測) * 常見tinyML開發板(自帶攝影機) * 常見具視覺TinyML開發板 * TinyML MCU 等級相關開發板 * Arm Cortex-M55/M85 + Ethos-U55/U65 * 單晶片推論速度與耗能比較 * [各式TinyML視覺開發板效能比較](https://www.hackster.io/mjrobot/computer-vision-at-the-edge-with-grove-vision-ai-module-v2-0003c7) 1.4. TinyML 常見開發工具 * TinyML 應用開發流程 * 資料集建置工具 * [Edge Impulse Data acquisition - Data Source](https://docs.edgeimpulse.com/docs/edge-impulse-studio/data-acquisition/data-sources) * [SensiML Data Capture Lab](https://sensiml.com/documentation/data-studio/data-collection-overview.html) * [CVAT](https://www.cvat.ai/) * [Roboflow](https://roboflow.com/) * [SenseCraft AI](https://sensecraft.seeed.cc/ai/) * 開發框架與工具 (MCU) * TinyML 開發平台與技術供應商 * 開發框架 (CPU / GPU / NPU / MCU) * [TensorFlow](https://www.tensorflow.org/?hl=zh-tw) * [TensorFlow Lite (TFL)](https://www.tensorflow.org/lite/guide?hl=zh-tw) * [TensorFlow Lite for Microcontrollers (TFLM, TFLu)](https://www.tensorflow.org/lite/microcontrollers?hl=zh-tw) * [PyTorch](https://pytorch.org/) * [PyTorch Mobile](https://pytorch.org/mobile/home/) * [ExecuTorch](https://pytorch.org/executorch-overview) * [uTensor](https://github.com/uTensor/uTensor) * [microTVM](https://tvm.hyper.ai/docs/topic/microtvm/) * [AIfES](https://github.com/Fraunhofer-IMS/AIfES_for_Arduino) * MCU硬體廠商收購TinyML軟體開發商 * [ML Commons (MLPerf) Benchmarks:Tiny](https://mlcommons.org/benchmarks/inference-tiny/) ## 參考文獻 [1] 許哲豪,NTUST Edge AI 人工智慧與邊緣運算實務 https://omnixri.blogspot.com/p/ntust-edge-ai.html [2] 許哲豪,OmniXRI's Edge AI & TinyML 小學堂 Youtube 直播課程總結 https://omnixri.blogspot.com/2024/06/omnixris-edge-ai-tinyml-youtube.html [3] 許哲豪,OmniXRI系列發文─TinyML(MCU AI)系列 https://hackmd.io/1PK1URhIQ7GutcWgpgsWbg#TinyMLMCU-AI%E7%B3%BB%E5%88%97 ## 延伸閱讀 * [Edge Impulse Studio Document](https://docs.edgeimpulse.com/docs) * [Seeed Studio SenseCraft AI](https://sensecraft.seeed.cc/ai/) * [Google TensorFlow Lite](https://www.tensorflow.org/lite/guide?hl=zh-tw) * [TensorFlow Lite for Microcontrollers](https://www.tensorflow.org/lite/microcontrollers?hl=zh-tw) * [Roboflow](https://roboflow.com/) * [arm Cortex-M55 Processor](https://developer.arm.com/processors/cortex-m55) * [arm Ethos-U55 MicroNPU](https://developer.arm.com/Processors/Ethos-U55) * [arm Common Microcontroller Software Interface Standard (CMSIS) V6 Document](https://github.com/ARM-software/CMSIS_6) * [Github - HimaxWiseEyePlus / Seeed_Grove_Vision_AI_Module_V2](https://github.com/HimaxWiseEyePlus/Seeed_Grove_Vision_AI_Module_V2) ## 教學資源 OmniXRI 系列文章: https://omnixri.blogspot.com/p/blog-page_19.html OmniXRI Youtube 教學影片頻道: https://www.youtube.com/@omnixri1784/videos OmniXRI Github 課程簡報及相關範例: https://github.com/OmniXRI/TinyML_Course_2025 --- **註:本課程非學校正式課程,現僅有老師一人,沒有教學助理可幫忙,如操作上有相關問題,請於[Youtube](https://www.youtube.com/@omnixri1784/featured), [FB Group](https://www.facebook.com/groups/edgeaitw), [Blogger](https://omnixri.blogspot.com/), [Medium](https://omnixri.medium.com/), [Hackmd](https://hackmd.io/@OmniXRI-Jack), [Github](https://github.com/OmniXRI) 各討論區中留言,老師會儘量協助,如有服務不週之處尚請見諒。**
×
Sign in
Email
Password
Forgot password
or
Sign in via Google
Sign in via Facebook
Sign in via X(Twitter)
Sign in via GitHub
Sign in via Dropbox
Sign in with Wallet
Wallet (
)
Connect another wallet
Continue with a different method
New to HackMD?
Sign up
By signing in, you agree to our
terms of service
.