Vision Navigator: A Smart and Intelligent Obstacle Recognition Model for Visually Impaired Users

Table 3

Accuracy rate analysis of observed obstacles with varying obstacle detection models in outdoor settings.

Environment	Model	Obstacle	Accuracy (%)	Mean accuracy (%)
Outdoor	Retina Net	Bus-1	82.1	84.76
		Human-1	70.2
		Car-1	91.6
		Tree	90.5
		Cow	89.4
Outdoor	Yolo Tiny	Car-2	83.6	81.32
		Human-2	81.1
		Human-3	82.2
		Truck	78.5
		Dog	81.2
Outdoor	R-CNN	Human-4	86.3	86.64
		Cycle	90.2
		Human-5	88.1
		Bus-2	88.5
		Human-6	80.1
Outdoor	SSD-RNN	Human-7	85.6	87.68
		Car-3	87.2
		Human-8	88.9
		Shop	90.1
		Human-9	86.8