train own dataset about keras-yolo2 HOT 4 CLOSED

First you need xml annotations files, you can generate them using LABELIMG.
As zenineasa said, just ignore the load pretrained weights section and modify:

LABELS = [...]

COLORS = [...]

CLASS = 3

# Layer 9
model.add(Conv2D(BOX * (4 + 1 + CLASS), (1, 1), strides=(1, 1), kernel_initializer='he_normal'))
model.add(Activation('linear'))
model.add(Reshape((GRID_H, GRID_W, BOX, 4 + 1 + CLASS)))

from keras-yolo2.

Just ignore the load pretrained weights section. It should work, i guess...

from keras-yolo2.

@zenineasa thanks ignore the weight.
and it work.

@Jonas1312 thanks!! I tried your solution and it work.
now I training.

from keras-yolo2.

@Heidisnaps @zenineasa @Jonas1312 you guys are awesome. I will fix layer 9 accordingly.

from keras-yolo2.