📂
JackyChen的精神時光屋
  • About
  • iPlayground
    • iPlayground 2020
      • iPlayground submit 2020
    • iPlayground 2019
      • Untitled
      • iPlayground submit 2019
  • WWDC
    • 2020 WWDC
    • 2019 WWDC
    • 2018 WWDC
      • What's New in Testing
      • BusinessChat
    • 2016 WWDC
      • What's New in the Apple Push Notification Service
  • AR/VR
    • ARKit plugin at Unity
    • ARKit
    • AR/VR 實習作品分享
    • Google Blocks
  • CI/CD
    • Continous Integretion for Unity
    • 拯救地球大作戰-自動化設定注意事項
    • Provisioning Profile 自動化更新
    • Make ipa file with personal team of code sign
    • Xcode11 版號問題
  • Test
    • Cucumberish
    • XCUITest
    • Design Patterns in XCUITest
    • Unit Test
  • User Interface
    • IBDesignable 和 IBInspectable
    • iOS 使用貝塞爾曲線繪製路徑
    • UIStatusBarStyle
    • iOS Devices Specification
    • Vector Image
    • Launch Screen
    • Haptic Feedback
    • Good Works for Storyboard
    • Cell 展開收合效果
    • ScrollView
    • Swift lazy
    • Lottie
  • Foundation
    • Adding a Custom Font to Your App
    • WKWebView how to work with javascript
    • Global Central Dispatch
    • HealthKit
    • Error Handling
    • Debug with LLDB
    • Application Lifecycle
    • Swift Codable
    • Push Notifications
    • AVFoudation
  • Others
    • C語言指標概念
    • UnsafePointer(Swift)
    • iOS News Reference
    • Blender
    • Free Web Server
    • Firebase
    • Firebase migration
    • GraphQL
    • Ruby
    • zsh command line
    • visudo
  • Security
    • 課程:App資安規劃與實作
    • KeyChain
    • iOS反組譯程序
    • Arxan
  • Git
    • Git
    • xcodeproj 合併衝突
    • Pull Request
  • Machine Learning
    • CoreML
    • Vision
    • Turi Create
  • 待分類
    • ABI Stability and More
    • Mirror
    • Carthage
    • SwiftUI
    • MVVM
    • OpenSSL
    • USDZ Convert
    • Nexus repository and gitlfs
Powered by GitBook
On this page
  • Python 環境設定
  • Turi Create

Was this helpful?

  1. Machine Learning

Turi Create

PreviousVisionNextABI Stability and More

Last updated 5 years ago

Was this helpful?

Python 環境設定

    • 使用 turi create 前,先設定好 Jupyter Notebook 環境

Turi Create

    • model 將識別目標物的 label 、大小及坐標

      [{'coordinates': {'height': 104, 'width': 110, 'x': 115, 'y': 216},
      'label': 'ball'},
      {'coordinates': {'height': 106, 'width': 110, 'x': 188, 'y': 254},
      'label': 'ball'},
      {'coordinates': {'height': 164, 'width': 131, 'x': 374, 'y': 169},
      'label': 'cup'}]
  • 利用合成的技巧,用少量的素材產生大量的圖資供 model 訓練用

    •   from PIL import Image
      
        # 讀取圖像 
        im = Image.open("lenna.jpg"
        im.show()
      
        #  放為 128x128
        im_resized = im.resize((128, 128))
        im_resized.show()
      
        # 旋轉
        im_rotate = im.rotate(45) 
        im_rotate.show()
      
        # 翻轉
        out = im.transpose(Image.FLIP_LEFT_RIGHT)
        out = im.transpose(Image.FLIP_TOP_BOTTOM)
        out = im.transpose(Image.ROTATE_90)
        out = im.transpose(Image.ROTATE_180)
        out = im.transpose(Image.ROTATE_270)
  • 直接下載圖的好用工具

在 Mac 上架設 Jupyter Notebook 環境
轉換 Jupyter(.ipynb) to Python(.py)
A Guide to Turi Create - WWDC2018
WWDC18, Session-712 Turi Create 重點整理
使用 Turi Create 製作圖案辨識的 App - Core ML and Vision
Object Detection
Python : 用 PIL 做影像處理(合成)
Python: 圖像的縮放、旋轉與翻轉
Google Images Download