Object Detection
Purpose of the Node
The Object Detection node evaluates ONNX-based object detection models for images. It takes an input image, an ONNX model session, and confidence and IoU thresholds as inputs, and returns bounding box predictions.
Pins
Pin Name | Pin Description | Pin Type | Value Type |
---|---|---|---|
Start | Initiate Execution | Execution | Execution |
Model | ONNX Model Session | Struct | NodeOnnxSession |
Image | Image Object | Struct | NodeImage |
Conf | Confidence Threshold | Float | Normal |
IoU | Intersection Over Union Threshold for NMS | Float | Normal |
Max | Maximum Number of Detections | Integer | Normal |
End | Done with the Execution | Execution | Execution |
Boxes | Bounding Box Predictions | Struct | BoundingBox (Array) |