How to get the confidence

[TX-Leo]

@warmshao @richardrl @BaldPulse

Hi guys, I'm wondering if there is any way to get the confidence of the detection result. Thanks!

[BaldPulse]

This implementation uses a YOLO model for detection.

WiLoR-mini/wilor_mini/pipelines/wilor_hand_pose3d_estimation_pipeline.py

Lines 74 to 81 in ebec42f

	self.logger.info(f"loading Yolo hand detection model >>> ")
	self.hand_detector = YOLO(yolo_model_path)
	self.hand_detector.to(self.device)
	@torch.no_grad()
	def predict(self, image, **kwargs):
	self.logger.info("start hand detection >>> ")
	detections = self.hand_detector(image, conf=kwargs.get("hand_conf", 0.3), verbose=self.verbose)[0]

The outputs of the detector follows ultralytics yolo output structure, which will contain confidence values.