Classify Image (Ultralytics YOLO)

How it fits

Typical backends

Caveats

• I/O contractA single ranked list is built per frame and every output arm is a projection of it, so all arms agree class-for-class; the arm is fixed by which output type the downstream connection resolves, not by a parameter. The arms are Classification (top-1), [Classification] (ranked), String (top-1 label), ([String], [Double]) (parallel index-aligned label and score arrays), [DetectedClass] (ranked id/confidence), UInt64 (top-1 class id), and [Double] (ranked confidence scores, descending).
• Hard constraintInference uses a local YOLO classification weights file bundled with the component; the label set is the checkpoint's own class names, and a class with no recorded name falls back to its stringified id.
• Parameter interactionThe confidence floor is applied before the top selection, so classes below it are dropped first and the kept-count then bounds the remaining ranked list.
• FallbackA requested GPU collapses to CPU when no CUDA device is visible, where per-frame latency rises.
• AccuracyThe square inference size rescales each frame, so very wide or tall inputs are distorted before classification.