I will build a costum computer vision model using roboflow

To begin, I need a clear description of what you want to detect or monitor. If you already have images or videos, you can provide them. If not, I can guide you on how to collect the right data for Roboflow.

Yes! I can handle the entire annotation (labeling) process in Roboflow using bounding boxes or polygons. If you have raw data, I will organize, tag, and augment it to ensure the model achieves high accuracy and precision.

I primarily work with YOLO (v8 through v11) for object detection and instance segmentation due to their high speed and accuracy in industrial environments. However, I can adapt the architecture based on your specific needs, whether for cloud deployment or edge devices like NVIDIA Jetson.

Absolutely. Being an Industrial Engineer, I specialize in connecting AI results with business metrics. I can export detection data to Power BI or RStudio to create real-time KPIs, allowing you to track production quality and efficiency (OEE) automatically.

During the preprocessing phase in Roboflow, I apply advanced "Augmentation" techniques (adjusting brightness, noise, and blur). This ensures the model is robust enough to maintain high performance even in challenging industrial environments with varying light or dust.