I think this code doesn't have data augmentation part while the original darknet has an "extensive data augmentation" as written in their paper.
Should I do it on my own?
def preprocess(self, im, allobj = None):
"""
Takes an image, return it as a numpy tensor that is readily
to be fed into tfnet. If there is an accompanied annotation (allobj),
meaning this preprocessing is serving the train process, then this
image will be transformed with random noise to augment training data,
using scale, translation, flipping and recolor. The accompanied
parsed annotation (allobj) will also be modified accordingly.
"""
darkflow/net/yolo/predict.py line 49.
What do you think does it mean?
@janchk Ok, thanks.
@janchk, This data augmentation code doesn't seem to appear in yolov2 folder? Could you throw some light on that?
@kmsravindra, I was also interested in this question, and actually this data augmentation is present in yolov2. The function (preprocess()) is just imported from yolo folder and called in _batch() function in darkflow/darkflow/net/yolov2/data.py line 27.
Most helpful comment
@kmsravindra, I was also interested in this question, and actually this data augmentation is present in yolov2. The function (preprocess()) is just imported from yolo folder and called in _batch() function in darkflow/darkflow/net/yolov2/data.py line 27.