Nathan Lam
Software Engineer / Fresh Graduate
Passionate to make great product with teams.
About Me
I first got my hands on programming in the Information and Communications Technology subject during high school, when I determined to go into Computer Science subject in the University.
During my university years, I had the opportunity to delve deeper into the world of coding and explore how the latest technology can enhance our lives. Beyond honing my technical abilities, one of the most valuable experiences I gained at university was collaborating with teams to create larger-scale products.
My projects include the research of a Smart Traffic Light System aimed at enhancing traffic efficiency through advanced algorithms and IoT technology. Also, I have consistently stayed abreast of the latest advancements in the field, as my final year project centers around the creation of Artificial Intelligence Generated Content.
Education
- The Hong Kong Polytechnic University
2019 - 2024
Finished my study as a full-time student at the University. Proficient in various skills through lectures, assignments, and projects.
- National Taiwan University
2019 Fall semester
Visited the National Taiwan University for a month. Broaden my horizon with different styles of teaching among universities and technology development in other places.
- Massachusetts Institute of Technology
2021 Summer
Participated in the Research Aboard Scheme lasted for 3 months. Designed group research project and met with the experts.
Experience
Aug 2021 - Dec 2021
Perform testing on Micro:bit extension using practical projects. Assit in developing an updated version of Micro:bit extension.
Jun 2021 - Aug 2021
Develop an IoT solution for Smart City, Smart Traffic Light. Develop an educational IoT project and provide training material for secondary school students.
My Work
This project aims to lower the cost of music production used as background audio in industrial scenarios. Also to explore the application of the Artificial Intelligence Generated Content (AIGC).
I handled all the way from the data sourcing from open-source music databases, data preparation, i.e. organizing the notes and chords as input data, to the machine learning model and layer selection. The final model successfully generated music with different moods and lengths suitable as background music.
Example of the AI-generated music with relaxing mood.
Another example of the AI-generated music with intense mood.
After finishing the model, some empirical validations have also been done on the model. Including the public assessment on the melody and musicality of the generated results through surveys, Putting the audio as background music of the game, like:
And the Ablation test on the model, which evaluates the performance of the model after the removal of a particular layer.
This project tried to explore better traffic light solutions to help with the common jaywalking in Hong Kong. Utilizing possible Internet of Things (IoT) resources to optimize the traffic light algorithms.
I was in charge of the maintenance of the sensors, the final calculation algorithms of the traffic light, the protocol of data transmission, and the distribution of the group work.
At the start, we proposed computer vision with a high-resolution camera for counting vehicles and pedestrians. Later, we found that weight sensors and communication with the smart car through network protocol are more effective solutions, considering the viability in the future. Here is a brief demonstration of the model.
First consideration of the algorithms is the number of predestrains, calculated by weight sensors.
Transcript
When people step on the weight sensor, representing waiting for the light, the pedestrian traffic light will turn green faster.
Second consideration of the algorithms is the number of vehicles, received by the protocol with the smart vehicles.
Transcript
When people cross the road, leaving the sensor below, and there are cars, the pedestrian traffic light will turn red faster.
Here is my interview with MIT News about the project.
https://news.mit.edu/2021/app-inventor-appathon-0930
Section: Cities of the future
...More when we get in torch!