Cecelia Soh
PROFESSIONAL SUMMARY
A Computer Engineering graduate from Nanyang Technological University, I’ve focused on Artificial Intelligence since June 2021. My expertise spans machine learning, computer vision, and image processing. I’ve gained practical research experience, especially in exploring face attribute editing with Generative Adversarial Networks (GANs) and designing custom data augmentation for face anti-spoofing. I also developed deep learning models for facial skin condition estimation during a project with Procter & Gamble. Committed to cutting-edge AI research and development, I’m excited to embark on my part-time Master of Computing in AI at the National University of Singapore, commencing August 2025.
EDUCATION
Bachelor of Engineering (Computer Engineering)
Aug 2019 – June 2023
Nanyang Technological University, Singapore
Graduated with Honours (Highest Distinction)
Specialising in Artificial Intelligence
PROFESSIONAL EXPERIENCE
March 2025 – Now
PC Partner Technology Pte. Limited
AI Engineer
- Designed and implemented localized AI agents using large language models (LLMs) for an embedded storage product, ensuring secure and privacy-preserving handling of consumer data.
- Integrated Retrieval-Augmented Generation (RAG), long-term memory, and tool-calling capabilities (e.g., search, code execution, API interaction) to support contextual, multi-step task execution, leveraging LangChain frameworks for modularity and performance optimization.
July 2023 – Dec 2024
Rapid-Rich Object Search Lab, NTU
Project Officer
- Engineered a deep learning framework for analysing facial skin conditions, which powered the development of a consumer app for Procter & Gamble, enhancing their skincare solutions.
- The developed framework pipeline included face detection, assessment point estimation, and skin assessment.
- Ensured seamless collaboration with stakeholders by aligning project deliverables with business goals, contributing to impactful and actionable outcomes.
Aug 2022 – May 2023
S-Lab, NTU
Final Year Project
Title: Generating Human Face by Generative Adversarial Networks
- Harnessed expertise in GAN models by mastering their conceptual framework, development, and applications.
- Developed a novel application integrating GANgealing for visual alignment with AttGAN for human facial attribute editing.
Jan 2022 – Dec 2022
Rapid-Rich Object Search Lab, NTU
Internship
Title: Face Anti-Spoofing Research Project
- Pioneered task-specific data augmentations for Face Anti-Spoofing (FAS), significantly enhancing the performance of ViT-based models through rigorous evaluation and comparison.
- Co-authored and published a peer-reviewed paper in International Journal of Computer Vision (IJCV)), showcasing the methodology and validating the impact of the proposed augmentations on FAS tasks.
PUBLICATIONS
Published:
- Rizhao Cai *, Cecelia Soh *, Zitong Yu, Haoliang Li, Wenhan Yang, and Alex C. Kot. “Towards Data-Centric Face Anti-spoofing: Improving Cross-Domain Generalization via Physics-Based Data Synthesis.” International Journal of Computer Vision (2024): 1–22.
Forthcoming:
- Cecelia Soh, Rizhao Cai, Monalisha Paul, Dennis Sng and Alex C. Kot. “AI-driven Remote Facial Skin Hydration and TEWL Assessment from Selfie Images: A Systematic Solution.” Machine Intelligence Research (2025).
(* indicates equal contribution)
HONORS & AWARDS
2023 - Graduated with Honours (Highest Distinction), B.Eng. Computer Engineering, NTU
2021 - Accenture Gold Medal – Book Prize, NTU SCSE, for academic excellence (Non-Graduating Year)
TECHNICAL SKILLS
- Programming Languages: Python, MATLAB, C/C++
- Machine Learning Frameworks: PyTorch, LangChain
- Data Processing: Numpy, Pandas, OpenCV
- Data Visualization: Matplotlib, SciPy
- Model Monitoring and Deployment: TensorBoard, MLFlow, Docker, AWS
LANGUAGES
- English
- Chinese
GitHub
LinkedIn
Google Scholar