Research Interests
current focus
Multimodal & VLMVision-language pretraining, cross-modal retrieval, visual reasoning, grounding
3D Vision & 3DGSDepth estimation, neural reconstruction, Gaussian splatting, spatial understanding
Generative AIDiffusion models, controllable generation, latent manipulation, inversion
Efficient LearningQuantization, distillation, pruning, LoRA, continual learning
Explainable AIAttention analysis, feature visualization, concept discovery, probing
Representation LearningSelf-supervised, contrastive learning, metric learning, embeddings
Selected Projects
production & research
All
Production
Vision
Generative
RAG Chatbot System — Enterprise Knowledge

Production RAG with FastAPI, LangChain, Pinecone, and OpenAI. Hybrid retrieval for structured + unstructured data, prompt orchestration, conversation memory, and AWS deployment. Built for scale with observability.

FastAPILangChainPineconeRAGAWSOpenAI
Vehicle Monitoring System — Real-time RTSP Analytics

End-to-end video analytics: YOLOv4 detection, ALPR, OCR, DeepSORT tracking, and re-identification. 40-45 FPS on GPU, 90% recognition accuracy, 4× model compression via quantization/distillation.

YOLOv4DeepSORTTensorRTSageMakerALPR
Text-to-Image Diffusion — From Scratch

Custom DDPM implementation in PyTorch with mixed precision, gradient accumulation, cosine schedule, and DDIM sampling. Trained on custom datasets with thorough logging and evaluation.

PyTorchDiffusionDDPMCUDA
VQ-VAE Image Compression

VQGAN-based compression pipeline with learned codebook, latent storage, and 8-bit serialization for efficient archival. Perceptual loss + adversarial training for high fidelity.

VQ-VAEPyTorchCompression
Paper Implementations — Deep Dive Series

Clean, from-scratch PyTorch re-implementations of landmark papers: GANs, diffusion, segmentation, and knowledge distillation. Focus on architectural understanding and reproducibility.

ResearchPyTorchReproducibility
Core Skills
toolkit

Computer Vision

Detection, Tracking, Segmentation, Pose, Depth, OCR/ALPR, Video Analytics, 3DGS

AI / ML

PyTorch, TensorFlow, HuggingFace, RAG, LLMs, Diffusion, GANs, Transformers, VLM

Engineering

Python, C++, Docker, FastAPI, ONNX/TensorRT, AWS/GCP, Linux, MLOps

Certifications
selected
NUS ISS — Specialist Diploma in AI (DL, CV, NLP) · 2020
DeepLearning.AIDeep Learning Specialization
DeepLearning.AIGANs Specialization
DeepLearning.AITensorFlow Developer
DeepLearning.AITF Advanced Techniques
Technical Blogs & Community
Links