UROP Proceedings 2020-21

School of Engineering Department of Computer Science and Engineering 109 Introduction to Brain-Computer Interfaces Supervisor: CHATZOPOULOS Dimitrios / CSE Student: CHEUNG Sze Yuen / COMP Course: UROP1100, Fall A Brain-Computer Interface (BCI) is an interface which allows direct communication between brain and modern computers. Recent research has opened possibilities for BCIs focus on enhancing performance on healthy people. With the advancement of Deep Learning, it could largely alleviate the need of manually extraction of features from the signals. In this research we explored a generalized convolutional neural network for EEG signals processing on the dataset ‘Attempted arm and hand movements in persons with spinal cord injury’. Compared to the baseline, EEGNet performed similarly in terms of accuracies. The bp+LDA classification performed similarly also. Our result suggest that the rise of usage of Deep Learning could accelerate the exploration of EEG and BCIs. The code can be found at: https://github.com/Arcaneless/eegnet and https://github.com/kshatilov/eeg-aarhm-sci/blob/main/KEEGclf.py Introduction to Brain-Computer Interfaces Supervisor: CHATZOPOULOS Dimitrios / CSE Student: ZHU Chen / COMP Course: UROP1100, Spring Blockchain has become an extremely hot topic in recent years. When we study the architecture and functionality of blockchain, the smart contract is always an important topic. In this project, I studied the basic concept of blockchain, the principle of the popular blockchain Ethereum, building smart contracts using solidity and how we interact with smart contracts using web UI. My report will be a conclusion of what I have learned in this project. It will focus on two topics: the principle and advantages of blockchain and how to build and interact with smart contracts. Introduction to Brain-Computer Interfaces Supervisor: CHATZOPOULOS Dimitrios / CSE Student: LAI Kei Chit / MAEC Course: UROP1100, Summer Smart contracts are computer programs that are stored and executed on the blockchain. Although there are several limitations, smart contracts are increasingly popular in different areas because all the functions and states are stored on the blockchain without reliance on any intermediary. The decentralization, immutability, transparency and security differ them significantly from traditional computer programs. In this project, I have learned the characteristics and applications of smart contracts on the Ethereum blockchain through some readings. I have learned the Solidity programming language and have gained experience in developing and using some representative examples, such as bank, voting and auction, on Remix IDE.