mmersive virtual reality chemistry lab for Meta Quest 2 that enables safe experimental practice and interactive XR simulations of laboratory procedures.
0
19
A dark-studio, sketchbook-themed portfolio for a portrait/line artist. Same stack as the AI-engineer portfolio: Next.js (App Router), Tailwind, Framer Motion, lucide-react, react-icons, and a working nodemailer contact form. The big thing is the AI , image to 3d model generator, which uses a tencent cloud API Stable Diffusion model to generate 3D models from 2D images. The models are then exported to GLTF format and displayed in the portfolio using Three.js.
0
22
The system is built upon a dual-model hybrid architecture, balancing inference speed with spatial accuracy. The pipeline is heavily optimized to run on local hardware accelerators.
1. Object Detection (YOLOv8) Standardized on Ultralytics YOLOv8 for realtime bounding box generation. Fine-tuned to recognize critical indoor obstacles (doors, benches, people, structural barriers) while maintaining high FPS on edge hardware.
2. Metric Depth Estimation Utilizes Depth Anything V2 (Metric Indoor Small). Unlike relative depth models, this Vision Transformer (ViT) architecture outputs exact physical distances in meters, crucial for accurate collision calculations.
3. Edge Quantization (FP16) To preserve metric distance accuracy, the pipeline avoids INT8 quantization. Models are converted using FP16 (16-bit float) via ONNX/TFLite formats, doubling inference speed without degrading depth measurements.
4. Custom Python Pipeline A unified OpenCV-based tracking and inference loop. It manages frame buffering, selectively triggers the heavier depth model at optimized frame intervals, and computes the Collision Envelope mathematics.
0
25
Intelligent Q&A assistant for farmers, powered by retrieval-augmented generation.
Curated 100+ agricultural documents and generated 1,000+ semantic knowledge chunks
Embedded using Sentence Transformers and indexed in ChromaDB for vector retrieval
Integrated Groq-hosted Llama models for accurate, context-grounded responses
Built and deployed a publicly accessible Streamlit application
Optimized retrieval through chunk-size tuning and prompt engineering
Validated retrieval and response quality against a curated test question set
0
31