I trained the network without batching, hence the input dimension of graph is (H,W,C) (not even [1,H,W,C]).
But during inference, I need predictions for multiple images (batched inference).
How can we achieve this
Asked
Active
Viewed 102 times
0
ram
- 31
- 3
-
I don't think you can train without batching. Please show your code (the line with input dimension of (H,W,C)) – Andrey Feb 24 '21 at 05:28
-
Just figured out they are expanding the dimensions using expandims as the first op of the graph – ram Feb 24 '21 at 06:20