AI Playground
Run object detection on satellite and geospatial images
Process video files and image sequences with RTMDet, YOLO, RT-DETR, or RF-DETR
Real-time stream detection from camera feeds and live sources with RTMDet, YOLO, RT-DETR, or RF-DETR
Classify land cover types and terrain analysis
Detect objects inside relevant land cover zones (buildings in urban, boats in water, etc.)
Segment objects by point click, auto-center, or bounding box using SAM (local ONNX or server API)
Run YOLO object detection & segmentation locally in the browser via ONNX — no server needed
Choose YOLO, SAM3, or Grounding-SAM for interactive video detection and segmentation
Build synthetic training datasets with augmentation, SAM3 segmentation, and custom object classes
Start new model training session with your datasets
Evaluate model accuracy and performance metrics
Compare uploaded inference results against ground truth
Download detection results, trained models, and outputs
Automated data preparation and processing pipelines
Test and debug segmentation masks and annotations