budiryan/COMP4901J-CS231n

COMP4901J: Deep Learning in Computer Vision

COMP4901J / CS231n

COMP4901J: Deep Learning in Computer Vision (a HKUST course). The assignments (except assignment 4) are heavily copied or adapted from Stanford's CS231n (Convolutional Neural Networks for Visual Recognition).

Assignment 1: Image Classification, kNN, SVM, Softmax, Neural Network

• Q1: k-Nearest Neighbor classifier
• Q2: Training a Support Vector Machine
• Q3: Implement a Softmax classifier
• Q4: Two-Layer Neural Network
• Q5: Higher Level Representations: Image Features

Assignment 2: Fully-Connected Nets, Batch Normalization, Dropout, Convolutional Nets

• Q1: Fully-connected Neural Network
• Q2: Batch Normalization
• Q3: Dropout
• Q4: Convolutional Networks
• Q5: PyTorch / TensorFlow on CIFAR-10

Assignment 3: Image Captioning, Network Visualization, Style Transfer, Generative Adversarial Networks

• Q1: Image Captioning with Vanilla RNNs
• Q2: Image Captioning with LSTMs
• Q3: Network Visualization: Saliency maps, Class Visualization, and Fooling Images
• Q4: Style Transfer
• Q5: Generative Adversarial Networks

Assignment 4: Q-learning, Deep Q-Networks, Policy Gradient, Model Network

• Q1: Basic Q-learning algorithms
• Q2: World navigation with DQN
• Q3: CartPole with Policy Gradient
• Q4: Model-based Reinforcement Learning