Syllabus

Logistic, Introduction, History.

Image processing: image gradient, filter, convolution.

Classic edge/corner/line detection methods: Canny edge detection, Harris corner detection, line fitting.

Edge/line detection: Canny edge detection, line fitting(RANSAC, Hough transform).

Corner detection and feature descriptors: Harris corner detection, feature.

The outline of deep learning. Network architecture: single-layer neural network, Multi-Layer Perceptron (MLP).

Convolution layer and CNN. Conv layer vs. Fully-connected layer. Neural network training: weight initialization, loss, optimizer (gradient descent, SGD, Adam), learning rate.

Underfitting and overfitting: batch normalization, skip link, augmentation, regularization.

Classification task: KNN, SoftMax, cross entropy loss.

Classification task: receptive field, architecture(VGG, ResNet…).

Segmentation task: K-Means, upsampling, architecture (FCN, UNet).

Object detector (SSD, RCNN series, YOLO); 2D Instance Segmentation.

Camera model.

Guest Lecture on 3D Representations, Neural 3D Reconstruction & 3D Deep Learning, Dr. Li Yi.

Recurrent Neural Networks (RNN) and Long Short-Term Memory (LSTM).

Attention Mechanisms, Transformers, and Vision Transformers (ViT).

Generative Adversarial Networks (GAN) and Variational Autoencoders (VAE).

Guest Lecture on Vision Language Models (VLM) by Dr. Ruochen Xu

Diffusion Models, Flow Matching Models, Diffusion Transformer (DiT)