Documentation

Models

Overseer composes a focused catalogue of models. This is the short version; the full spec sheets live on the Models page.

Model Role Status
YOLO11 Object detection shipped
ByteTrack Multi-object tracking shipped
Depth Anything V2 Monocular depth shipped
Real-ESRGAN Super-resolution shipped
YOLO11-seg Segmentation beta
Pose Estimator Body keypoints beta
ReID + Gait Re-identification shipped
ViT Vehicle Classifier Vehicle attributes beta

Each model has a dedicated page covering purpose, inputs and outputs, performance, advantages and limitations, expected latency, hardware and memory needs, configuration keys and benchmark placeholders.

Full catalogue

See the Models page for every spec sheet, and Technology for how they fit together.