query related to cornerNet inference

[Sujith93]

I trained cornerNet on custom data

After runing the below line of code
boxes = gtf.Predict("/data/train/Image_046.jpg", vis_thresh=0.5, output_img="output.jpg")
from IPython.display import Image
Image(filename='output.jpg')

I tried printing boxes as I'm looking for predicted bounding boxes and its related scores, but from the output, it looks something like this and unable to interpret this.

boxes : {'Apple': array([[1.50363940e+03, 9.93703247e+02, 2.09030884e+03, 1.20912891e+03,
6.05484009e-01],
[1.49551221e+03, 9.82063232e+02, 3.78910303e+03, 1.13179028e+03,
3.35086226e-01],
......
[3.8891382e+02, 4.1944517e+03, 2.7675557e+03, 4.3056367e+03,
1.5905794e-03],
[4.4121234e+02, 2.8287498e+03, 3.7419502e+03, 3.0195723e+03,
1.2366427e-03]], dtype=float32), 'Banana': array([[4.3748615e+02, 1.3137083e+03, 1.4082772e+03, 1.4733477e+03,
7.0229608e-01],
[3.9888440e+02, 2.6156851e+03, 1.3096824e+03, 2.7784297e+03,
6.4267540e-01],
......
[1.6594436e+03, 7.6432733e+02, 2.0213037e+03, 1.4048514e+03,
1.9753652e-03]], dtype=float32)}

Could you please help me out how to get the bounding boxes and its related scores from this output.

[THEFASHIONGEEK]

    bboxes: A dictionary representing bounding boxes of different object
        categories, where the keys are the names of the categories and the
        values are the bounding boxes. The bounding boxes of category should be
        stored in a 2D NumPy array, where each row is a bounding box (x1, y1,
        x2, y2, score).

For example if we see your output, first key-value pair you have
{'Apple': array([[1.50363940e+03, 9.93703247e+02, 2.09030884e+03, 1.20912891e+03,
6.05484009e-01] ....}
here Apple is the class x1 =1.50363940e+03,y1= 9.93703247e+02,x2 = 2.09030884e+03, y2= 1.20912891e+03, score =6.05484009e-01.

[Sujith93]

yes, the values look confused, I got it.
Thanks.