State Of The Art Object Detection

Satellite image deep learning · Feb 4, 2026 · 30m · transcript

In this episode I sat down with Isaac to discuss RF-DETR, a new state-of-the-art family of real-time object detection and segmentation models from Roboflow. We cover the motivation for building models that are not just accurate but also fast, cost-efficient, and deployable across diverse hardware and data regimes, and why moving beyond fixed architectures is key to achieving that. Isaac explains how RF-DETR combines strong foundation backbones like DINOv2 with efficient neural architecture search to unlock novel speed–accuracy trade-offs, including dropping decoder layers and queries after…