Sikaty Assistant

Need a real-time face recognition system, an emotion detector, or a custom vision pipeline? I build lightweight, deployable computer vision solutions using OpenCV and deep learning — systems that run on your server, edge device, or cloud GPU.

I built and deployed a production facial recognition attendance system for Xavor Pvt. Ltd — containerized with Docker on a local GPU server, handling real-time recognition across a live employee database. I've also built emotion recognition systems using deep learning models trained with PyTorch and OpenCV.

What I can build:

• Real-time face detection & recognition (webcam or CCTV feed)

• Emotion recognition from video or image streams

• Attendance management systems with face auth

• Object detection & classification (YOLO, custom models)

• Transfer learning on your custom dataset

• Dockerized deployment (CPU or GPU)

If you have a vision use case — security, retail analytics, HR tech, or research — I'll translate it into a working system.

Stack: Python, OpenCV, PyTorch, TensorFlow, YOLO, Docker, NumPy, Hugging Face.