Here is Jianxu (Alan)

I am looking for a graduate program and research assistant position in ECE, CS, DS, AI, or any related fields in 2025FALL. Feel free to contact me if you are interested! Thanks!!!

Hi there! I’m currently a final year Computer Science undergraduate student at University of Nottingham, China Campus, advised by professor Jianfeng Ren in Artificial Intelligence and Optimization Group.

Before that, I had the privilege of working with professor Wei Tang at BINN, where I explored the integration of image processing algorithms like Gaussian Filtering with the U-Net based deep learning model to further refine the performance of segmenting retinal blood vessels from fundus images. I also spent a rewarding summer with professor Tianxiang Cui and other teammates, investigating the performance of using Quantum Reinforcement Learning to solve Electric Vehicle Charging problem.

My main research interest lies in Abstract Visual Reasoning, where I aim to test the ability of intelligent agents to apply acquired knowledge and experience to reason in new contexts. It has been shown that computer vision demonstrates amazing capabilities in areas such as image classification and object detection, while AVR tasks focus on identifying abstract relationships and patterns between shapes, emphasizing the reasoning and abstraction capabilities based on visual input. Given the current problems of abstract visual reasoning, I try to make further research in the following areas.

1. Model Generalization: Although deep learning methods perform well on specific AVR tasks, significant challenges remain when handling multiple types of problems. Models often tend to over-fit specific tasks and struggle to generalize to new ones.

2. Three-dimensional Visual reasoning: Extending AVR from 2D to 3D scenarios is particularly important for those fields that need to manipulate and understand complex objects (such as robotics and Autonomous driving).

3. Few-shot Learning: Algorithms are far less flexible than humans when it comes to concept learning and reasoning from a few examples. So I aim to develop models that can converge to an optimal state with minimal training data.

Outside of my academic pursuits, I enjoy photographing, playing table tennis, gaming and tracing sunrise. Please refer to my resume if you are interested!