05/2023 –
Developing "CosyWake" an innovative wearable app for addressing oversleeping issues. Researched and integrated a sleep phase classification model using the Machine Learning Long Short-Term Memory (LSTM) technique.
Implemented data feature extraction and analysis for accurate detection of light sleep phase. Created a user-friendly interface for Apple WatchOS using SwiftUI and MVVM architecture.
06/2023 – 08/2023
With a focus on optimizing GPU-accelerated kernels, I made a remarkable breakthrough in creating the fastest algorithm for computing SDOT.
By leveraging CUDA programming and innovative techniques, I successfully enhanced kernel efficiency and minimized data movement.
The project of my SDOT Kernel can be found at: https://github.com/yiliu-li/Optimized-Cuda-SDOT-Kernel-on-NVIDIA-Turing-GPUs.
05/2022 – 07/2022
Refraining from touching public facilities is one of a key to epidemic prevention and control. To address it, I made an automatic fingerbot, a wireless tiny mechanical arm that can operate a switch or button without any physical touch to the switch, with a sensor to detect a "pressing down" gesture.
I wrote the programme to the chip by Arduino, a C++-based coding platform that encloses plenty of Internet of Things (IoT)-related libraries.
09/2019 – 09/2019
Attended lectures about software and system development, which depicts the foundamental of the Distributed Operating System。
Participated in the Codelab activity and enjoyed the application development in Harmony OS.