📌 [January 2022 - September 2024]: Robust Active Inference

• Develop all pipelines and algorithms for an active inference agent that tackles sparse reward partially obsevable Markov decision processes (POMDPs). These POMDP environments have continuous action space, image-based observation, sparse reward setting, and varied goal.
• Github Repo for Robust Active Inference work.
• Website of the project
• Video of the project
• Research project's preprint

📌 [May 2023 - Present]: Event-based eye tracking research

• Designing a spiking neural network variant of the recurrent state space model for tracking eye features from dynamic vision sensor's event stream.
• Report to Meta's staff.

📌 [January 2023 - Present]: Robotic research

• Building a generative model for a reinforcement learning agent with an active inference based decision making system. Manuscript in preparation.
• Building active inference agent for robotic tasks. Manuscript in preparation.
• Secure a $1,000 grant to purchase robotic equipment that supports the research.

📌 [April 2022 - November 2023]: Eye segmentation research

• Develop new eye segmentation training algorithm that minimize the domain shift across different eye segmentation datasets. Best paper award at ETRA 2024.
• Replicate RITNet segmentation network for eye segmentation.
• Github Repo for RITNet Replication.
• Github Repo for Deep Domain Adaptation published work.
• Report to Meta's staff.

📌 [Dec 2021 - Jan 2022]: Music Emotion Recognition Algorithm

• Research and compare deep learning models in the music emotion recognition task.
• Github Repo

📌 [Dec 2021 - Jan 2022]: YOLO Algorithm Replication

• Replicate all version of YOLO training and evaluation algorithm.
• Github Repo

📌 [May 2021 - August 2021]: Credit card OCR

• Research and implement the algorithms for OCR.
• Try out different methods: template matching, training with Tensorflow Object detection API, training YOLO, training darknet.
• Build transfer learning model.
• Github Repo

📌 [May 2021 - Sep 2021]: Faster R-CNN Research and Implementation

• Code and research Faster R-CNN algorithm.
• Github Repo
• Backlog

📌 [Apr 2021 - May 2021]: Shopee's Kaggle Competition

• This project is created for academic purpose and also available publicly. More information about the competition can be found here.
• Github Repo
• Backlog

📌 [Feb 2021 - Apr 2021]: Chest X RAY Abnormalities Detection

• This project is created for academic purpose and also available publicly. More information about the competition can be found here.
• Github Repo
• Backlog

📌 [Jan 2021 - Feb 2021]: Atari Deep Q-learning network

• Research and implement reinforcement learning algorithms to play atari game.
• Implement deep Q-leanring method with variances. Implement rainbow methods which includes multiple techniques integrated in the deep Q-learning pipeline.
• Github Repo

📌 [Jan 2021 - Feb 2021]: Open-source contribution to topdup project

• Research and implement FAISS algorithm to search vectorized text.
• Github Repo

📌 [Dec 2020 - Jan 2021]: Virtual Try-on project

• This repo is the composition of multiple applications that can derive the generation of try-on image given the input image of the person and the image of the clothing.
• Github Repo

📌 [Nov 2020 - March 2021]: Open-source contribution to mlpack library

• Contribute to mlpack by fixing bugs, creating new features, and answering questions regarding issues.
• Github Repo
• Copyright: My name is on there.

📌 [Oct 2020 - Nov 2020]: Ebay ML Challenge

• Modify the method of Birch Tree, not using a branching factor but rather the fanout of each node is the number of generated clustering features.
• Github Repo
• Report

📌 [Aug 2020 - Dec 2020]: AI course work and research

• This repository is the collection of AI Idea implementation and research. This repository also contains small-scale to large-scale projects that are related to AI, ML, RL, and Wavelet Transformation.
• Github Repo

📌 [May 2020 - Sep 2020]: Algorithm of Reinforcement Learning for Imperfect Information Card Game: Gin Rummy

• Research and implement AI bot that optimize the 2-player card game Gin Rummy.
• Paper was accepted at AAAI 2021. Link to paper
• Github Repo

📌 [January 2023 - May 2023]:

• Build a tensor computing library from scratch with GPU acceleration that operates on top of numpy.
• Same idea with JAX
• Github Repo

📌 [March 2021 - August 2021]:

• Build mobilie application with React Native which warns users of early forest fire.
• Google Play Link
• Appstore Link (will be updated in the future)

📌 [Jan 2021 - Feb 2021]: Open-source contribution to topdup project

• Write backend server with nodejs, express, javascript.
• Perform dataset operations.
• Github Repo