In deploy.prototxt, the input shape is defined as
shape: { dim: 10 dim: 3 dim: 224 dim: 224 }
I know the meaning of the last 3 dims. They define the channel, height, and weight of an image. But what is the meaning of the first dim? Why is it 10?
I know that the first dim is the batchsize in train_val.prototxt. I set the first dim to be 10, and predict one image each time. If its meaning is still batchsize, why do I get no error or warning?
Thanks!