VRDL Lab1: Image Classification

This is homework 1 in NYCU Selected Topics in Visual Recognition Deep Learning class.

Hardware

The following specs were used to create the original solution.

• Google Colab
• PC
  • Intel(R) Core(TM) i7-7700 CPU @ 3.60GHz
  • NVIDIA RTX 1060 6G

Environment

• Python 3.0+
• Pytorch
• Matplotlib
• Urllib
• PIL

Dataset Preparation

https://drive.google.com/file/d/1HE5agKoWtAXCS32aYs5owfLq-UbvsnUW/view?usp=sharing

Requirement

• After downloading dataset, unzip the file named 'birds'.
• Unzip the training images and testing images in the birds folder.

Train and Valid

The code not only trains, but also valid and test the model. You can run the model by following:

python3 main.py

Inference

The code not only trains, but also valid and test the model. You can run the model by following:

python3 inference.py