This course will be about modern artificial intelligence and machine learning. It will start with a general overview of five fundamental neural networks (MLP, CNN, RNN, Autoencoder and Transformer) and followed by discussions of text, image and video generation techniques.
Tentatively, topics may include the following:
* CNN, RCNN, Fast RCNN, Faster RCNN, Yolo, Mask RCNN,
* RNN and LSTM,
* Attention and Transformer, Word2Vec, Elmo, Bert, and GPT,
* PixelCNN/RNN, Autoencoder, Variational Autoencoder, and VQ VAE,
* Diffusion models (Dall-e and Glide),
* Flow-based generation, Nice, Real NVP, and Glow,
* Generative and Adversarial Networks.
Topics may be modified depending on the interests of the participants and current R&D directions.
The minimum preparation is CMPSC 130B. CMPSC 165B is highly recommended but CMPSC 165A can serve as a substitute.