jackylu
jackylu
@linzhi123, could you provide your train data (1,000 pictures to train),thanks.
me too , I also like to study it.
xxxxx\CNN-Numpy-master\tensor\util.py", line 13, in initializer weights_scale = math.sqrt(reduce(lambda x, y: x * y, shape) / shape[-1]) NameError: name 'reduce' is not defined 请问这个reduce是哪里的函数啊
lenet_layers.py 和 lenet_tensor.py 哪个更新啊？区别在哪里啊？我lenet_layers.py 可以跑了。
I also find the four vertices locations is not accurate. but the image's number of this kind problem number is very small for the total data set.
I want to knwo whether we can use Bounding box coordinates to location the plate as training set.becuase I found for some cases, e.g. under tilt case, the four vertices...