| CARVIEW |
Jing-En Huang | 黃靖恩
Mathematics student passionate about artificial intelligence and computer vision. Seeking to leverage strong mathematical expertise and AI skills to develop innovative real-world solutions. Experienced in creating real-time automated systems for sports analytics, leading successful teams in AI competitions, and actively sharing expertise through numerous technical talks and presentations at international and national platforms. Currently, I’m interested in leveraging Diffusion models to create synthetic datasets that aid in solving real-world tasks.
Timeline
Jul. 2025 - Present
Master Student
National Yang Ming Chiao Tung University
Institute of Data Science and Engineering
- Comp Photo Lab
- Supervised by Prof. Yu-Lun Liu and co-advised by Prof. Jun-Cheng Chen from Academia Sinica
Dec. 2024 - Apr. 2025
Army
Ministry of National Defense
Mandatory Military Service
- Served a four-month mandatory military service in the Army at Jinliujie Barrack, Yilan.
Jan. 2024 - Present
Research Assistant
Academia Sinica
Research Center for Information Technology Innovation
- AIIU Lab
- Supervised by Prof. Jun-Cheng Chen and co-advised by Prof. Chih-Yu Wang
Jun. 2022 - Jun. 2023
Research Assistant
National Center for Theoretical Sciences
Undergraduate Research Program
- Supervised by Prof. Mei-Heng Yueh
- Published Jing-En Huang, Jia-Wei Liao, Ku-Te Lin, Yu-Ju Tsai, Mei-Heng Yueh. An Improved Variational Method for Image Denoising
May. 2022 - Jun. 2024
Research Assistant
National Taiwan Normal University
Research Combining Mixed Reality Display and the Field of Artificial Intelligencem
- Supervised by Prof. Jann-Long Chern
Jul. 2021 - Jun. 2023
Research Assistant
National Taiwan Normal University
Practical Applications of Computational Geometry in Three-Dimensional Imaging
- Supervised by Prof. Mei-Heng Yueh
Jul. 2020 - Jun. 2024
Bachelor Student
National Taiwan Normal University
Department of Mathematics
- Teaching Assistant of Computer Programming
- Related Courses:
- Mathematics: Numerical Analysis, Linear Programming, Statistics, Probability
- Computer Science: Deep Learning for Computer Vision, 3D Computer Vision with Deep Learning Applications, Data Science and Computer Programming, Computer Programming and Data Analysis in Sports
Publications
- Jing-En Huang, Jia-Wei Liao, Ku-Te Lin, Yu-Ju Tsai, Mei-Heng Yueh. (arXiv preprint 2024) An Improved Variational Method for Image Denoising.
Projects
[Independent Development] | Aug. 2022 - Jun. 2024
A Real‑Time Automatic System of Tracking and Analyzing Basketball Matches based on Computer Vision and Deep Learning Model
- Engineered a real-time system for tracking and analyzing basketball matches via computer vision.
- Developed an end-to-end system, reducing positioning error by 35% compared to infrared devices.
- Implemented the system in the NTNU gymnasium for coaches to analyze basketball matches.
[Project Leader] | Nov. 2023 - Dec. 2023
Enhanced Real-World VQA: A Selective-Based Approach
[Project Leader] | Jul. 2022 - Jun. 2023
An Improved Variational Method for Image Denoising
- Proposed a Total Variation (TV) model, improving denoising performance by 60% compared to classic TV methods.
- Conducted theoretical proof for convergence of the model.
- Extended the method to surface denoising through surface parametrization.
Mixed-Norm model stands for our improved variational method. PPS stands for product of PSNR and SSIM.
[Project Leader] | Feb. 2022 - Jun. 2022
An Improved End-to-end Framework for Image Stitching
- Implemented an image stitching framework, achieving 40% faster processing than Photoshop.
- Proposed a mismatch detecting algorithm, enhancing the stitching accuracy of the SIFT algorithm.
- Developed a Poisson-based smoothing method for integrating images with severe chromatic aberration.
- Introduced a conformal reshaping method to refine results and ensure image data integrity.
[Independent Development] | Nov. 2022 - Dec. 2022
A Small Object Detection Framework for Unmanned Aerial Vehicles Images
- Proposed a Small Object Augmentation (SOA) algorithm to improve the detection model’s ability to recognize small objects.
- Achieved a 30% improvement in small object recognition compared to models without the SOA algorithm.
[Independent Development] | Apr. 2023 - May. 2023
An AI Referee of Badminton Matches
- Developed a pipeline for analyzing single-camera perspective broadcasts of badminton matches.
[Project Leader] | Oct. 2022 - Dec. 2022
Crop Image Recognition
- Received the Honorable Mention Award in the AI CUP competition.
[Independent Development] | May. 2022 - Jun. 2022
Spread Through Air Spaces Tumor Detection
- Received the Honorable Mention Award in the AI CUP competition.
Honors & Awards
- Excellent Poster Paper Award
- Most Popular Award
- Excellent Presentation Award
- Honorable Mention Award
- Honorable Mention Award
- Most Popular Award
- Excellent Presentation Award
- Excellent Poster Paper Award
- Honorable Mention Award
- Excellent Presentation Award
- 2021 Taiwan Society for Industrial and Applied Mathematics
- 2021 Taiwan Society for Industrial and Applied Mathematics
- The 11th Japan‑Taiwan Joint Workshop in Applied Mathematics
- AI CUP 2021 Fall UAV Object Detection Competition
- AI CUP 2021 Fall Crop Image Recognition Competition
- 2022 Taiwan Society for Industrial and Applied Mathematics
- The 10th Japan‑Taiwan Joint Workshop in Applied Mathematics
- 2020 Taiwan Society for Industrial and Applied Mathematics
- AI CUP 2020 Spring STAS Object Detection Competition
- The 9th Japan‑Taiwan Joint Workshop in Applied Mathematics
Talks
- TWSIAM 2024 Poster Presentation | NCHU, Taichung | May. 2024
Enhanced Real-World VQA: A Selective-Based Approach - SITCON 2024 | Academia Sinica, Taipei | Mar. 2024
Enhanced Real-World VQA: A Selective-Based Approach - The 14th Japan-Taiwan Joint Workshop for Young Scholars in Applied Mathematics | Meiji, Tokyo | Feb. 2024
Multi-Camera Multi-People Tracking on Basketball Matches Based on Deep Learning Models - sciwork Conference 2023 | NYCU, Hsinchu | Dec. 2023 How Can We “Perfectly and Rapidly” Stitch Images?
- NCTS Undergraduate Research Program Presentation | NCTS, Taipei | Jun. 2023
Mixed-Norm Total Variation Denoising Model with Application to Surface Imaging - TWSIAM 2023 Poster Presentation | NTNU, Taipei | May. 2023
An AI Referee of Badminton Matches - NCTS Student Workshop on Scientific Computing | NCU, Taoyuan | Apr. 2023
An Improved Variational Method for Image Denoising - The 13th Japan-Taiwan Joint Workshop for Young Scholars in Applied Mathematics | NTU, Taipei | Feb. 2023
A Real-Time Automatic System of Tracking Basketball Matches Based on Deep Learning Models - NCTS Undergraduate Summer Research Program Final Presentation | NCTS, Taipei | Jul. 2022
Numerical Methods for Image and Geometry Processing - TWSIAM 2022 Poster Presentation | NYCU, Hsinchu | Jul. 2022
Real-Time Automatic Tracking of Basketball and Players by Computer Vision Based on YOLOv5, ByteTrack, FastReID and EasyOCR Deep Learning Models - The 12th Japan-Taiwan Joint Workshop for Young Scholars in Applied Mathematics | Meiji, Tokyo | Feb. 2022
A Study on the Image Stitching Technology - NCTS Student Mini-Symposium in Applied Mathematics NTHU, Hsinchu | Feb. 2022
A Study on the Image Stitching Technology
Certifications
- Programming Teaching Assistant of Department of Mathematics
- National Center for Theoretical Sciences Mathematics Division Undergraduate Research Program
- National Center for Theoretical Sciences Mathematics Division Undergraduate Summer Research Program
- Taipei Municipal Cheng-Gong High School Science and Math Talented Class
- Go Professional Dan Certificate
Programming Skills
- Languages: Python (Advanced), MATLAB (Advanced), C++ (Advanced), C (Intermediate), Java for Android (Basic), C# (Basic), PHP (Basic), LaTeX (Intermediate)
- Machine Learning Frameworks: PyTorch (Advanced), Scikit-learn (Intermediate)
- Data Analysis Libraries: NumPy (Advanced), Pandas (Intermediate)
- Version Cintrol: Git (Advanced)
- Joke: My colleagues like to respectfully call me the “Conda Environment Engineer” because I excel at successfully setting up conda environments in a short amount of time, which they often struggle with for days.