/CS231n

Spring 2017

Primary LanguageJupyter Notebook

CS231n Convolutional Neural Networks for Visual Recognition

This repository contains my solution for the cs231n of Stanford University.
If you have any questions, feel free to contact me via e-mail(ding@ivanpp.me).

Assignment1

  • Q1: k-Nearest Neighbor classifier (20 points)
  • Q2: Training a Support Vector Machine (25 points)
  • Q3: Implement a Softmax classifier (20 points)
  • Q4: Two-Layer Neural Network (25 points)
  • Bonus: Implement some extra techniques on Two-Layer Neural Network (5 points)
    see: ./Assignment1/cs231n/classifiers/neural_net.py and ./Assignment1/two_layer_net.ipynb
    I implement the dropconnect in neural_net.py and get 57.1% test accuracy in two_layer_net.ipynb
  • Q5: Higher Level Representations: Image Features (10 points)
  • Bonus: Design your own features!
  • Q6: Cool Bonus: Do something extra! (+10 points)

Assignment2

  • Q1: Fully-connected Neural Network (25 points)
  • Q2: Batch Normalization (25 points)
  • Bonus: Batch Normalization: alternative backward (3 points)
    see: batchnorm_backward_alt() function in ./Assignment2/cs231n/layers.py
  • Q3: Dropout (10 points)
  • Q4: Convolutional Networks (30 points)
  • Q5: PyTorch / Tensorflow on CIFAR-10 (10 points)
  • Bonus: Do something extra! (up to 10 points)

Assignment3

  • Q1: Image Captioning with Vanilla RNNs (25 points)
  • Q2: Image Captioning with LSTMs (30 points)
  • Bonus: Train a good captioning model!
  • Q3: Network Visualization: Saliency maps, Class Visualization, and Fooling Images (15 points)
  • Q4: Style Transfer (15 points)
  • Q5: Generative Adversarial Networks (15 points)
  • Bonus: WGAN-GP or something cool.