Darknet::Output_Object Struct Reference

This is used to help keep some state between calls to functions fill_network_boxes(), get_yolo_detections(), etc. More...

#include "darknet_internal.hpp"

Collaboration diagram for Darknet::Output_Object:

Public Attributes
int	i
	The entry index into the W x H output array for the given YOLO layer.
int	layer_index
	The layer index where this was found.
int	n
	What is "n"...the mask (anchor?) number?
int	obj_index
	The index into the YOLO output array – as obtained from yolo_entry_index() – which is used to get the objectness value. E.g., a value of "l.output[obj_index] == 0.999f" would indicate that there is an object at this location.

Detailed Description

This is used to help keep some state between calls to functions fill_network_boxes(), get_yolo_detections(), etc.

We use the cache to track objects within the output array, so we don't have to walk over the entire array every time we need to find all the objects and bounding boxes.

Since

2024-06-02

See also

yolo_num_detections_v3() where the cache is populated (in yolo_layer.cpp)

get_yolo_detections_v3() where the cache is accessed and converted to bounding boxes

make_network_boxes_v3()

Member Data Documentation

i

int Darknet::Output_Object::i

The entry index into the W x H output array for the given YOLO layer.

layer_index

int Darknet::Output_Object::layer_index

The layer index where this was found.

n

int Darknet::Output_Object::n

What is "n"...the mask (anchor?) number?

obj_index

int Darknet::Output_Object::obj_index

The index into the YOLO output array – as obtained from yolo_entry_index() – which is used to get the objectness value. E.g., a value of "l.output[obj_index] == 0.999f" would indicate that there is an object at this location.

---

The documentation for this struct was generated from the following file:

• darknet_internal.hpp