Mini Project 2 from BU EC601.
TensorFlow, Numpy need to be installed first.
This mini project aims to better understand the basic workflow of deep learning.
- Step1: DownloadImage
Using DownloadImage.py to download images from website by key words.
- Step2: Preprepocessing
Prepocessing images downloaded from step1, including reshape and rename.
- Step3: Add Label to Image
Making label list for every catogory.
- Step4: Build Neural Network
Building a CNN to classify images. The CNN model only has two convolution layers.
- Step5: Validate & Test
When training the model, if we set the learning rate to 0.01, the loss function will keep at 0.69 after 100 epochs. However, if we set the learning rate to 0.001, the loss function will decrease to nearly 0 after 100 epochs.