Boat Detection Model Creation

We are seeking a skilled freelancer to develop a boat detection model using video cameras. The ideal candidate will have experience in machine learning and video processing, and be able to create a model that accurately identifies and tracks bots in real-time. The project requires an intermediate level of proficiency and offers a flexible engagement with a part-time commitment. Hardware & Tech Stack: Daytime Camera: Dahua SD6CE445GB-HNR (Optical / PTZ RTSP stream) Nighttime Camera: Dahua TPC-SD2241 (Thermal RTSP stream) Deployment Edge Hardware: Raspberry Pi 4B Key Responsibilities: 1. Model Selection & Training: Fine-tune a lightweight object detection model (e.g., YOLOv8/v11-nano, MobileNet, or similar) specifically tailored for detecting small vessels (12m) on open water. 2. Dual-Stream RTSP Pipeline: Build a robust video ingestion pipeline handling Dahua RTSP streams (optical by day, thermal by night). 3. Edge Optimization: Quantize and optimize the model to achieve viable FPS and low latency on the Raspberry Pi 4B CPU. Deployment & Delivery: Provide clear documentation, deployment scripts, or Docker containers ready to run on the Raspberry Pi.