晶文摘

[人工智慧] 深度學習(DL)及人工智慧的應用實作

給晶新聞一個讚



深度學習(DL)及人工智慧的應用實作


作者: 夏肇毅

初稿: 20220819





電腦視覺 (Computer Vision)

科技> [人工智慧] CNN,影像分區塊與RNN


Convolutional Neural Network (CNN)

https://www.tensorflow.org/tutorials/images/cnn?hl=zh-tw

Import TensorFlow

Download and prepare the CIFAR10 dataset

Verify the data

Create the convolutional base

Add Dense layers on top

Compile and train the model

Evaluate the model


Basic classification: Classify images of clothing

https://www.tensorflow.org/tutorials/keras/classification?hl=zh-tw

Import the Fashion MNIST dataset

Explore the data

Preprocess the data

Build the model

Set up the layers

Compile the model

Train the model

Feed the model

Evaluate accuracy

Make predictions

Verify predictions

Use the trained model



影像處理 (Image Processing)

Generative

Neural style transfer

 https://www.tensorflow.org/tutorials/generative/style_transfer?hl=zh-tw

Deep Convolutional Generative Adversarial Network

 https://www.tensorflow.org/tutorials/generative/dcgan?hl=zh-tw

Intro to Autoencoders

 https://www.tensorflow.org/tutorials/generative/autoencoder?hl=zh-tw

Import TensorFlow and other libraries

Load the dataset

First example: Basic autoencoder

Second example: Image denoising

Define a convolutional autoencoder

Third example: Anomaly detection

Overview

Load ECG data

Build the model

Detect anomalies