Timeline for How can I reshape (16,) and (3,112,112) shape to the single (16,3,112,112)? See code below
Current License: CC BY-SA 4.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Aug 17, 2021 at 10:34 | review | Close votes | |||
| Aug 28, 2021 at 3:36 | |||||
| Aug 9, 2021 at 7:50 | comment | added | Tariq Hussain | Okay, Okay, I got you. | |
| Aug 9, 2021 at 7:50 | comment | added | Oxbowerce |
Like I said, use the numpy.stack equivalent for the library you are working with, which would be tf.stack for tensorflow and torch.stack for pytorch.
|
|
| Aug 9, 2021 at 7:37 | comment | added | Tariq Hussain | the each image is in tensor format and then the all images together are in list. I know this, but how can I make it all for tensor only so that I can only read it as [16,3,112,112]? | |
| Aug 9, 2021 at 7:34 | comment | added | Oxbowerce |
That is because you are appending tensors to a list, and since they are two different types at the top level the array is of shape (16,) whereas within each item the shape is (3, 112, 112). Instead of appending to the list you should use numpy.stack, or given that you are working with tensors, use the tensorflow/pytorch equivalent.
|
|
| Aug 9, 2021 at 7:31 | history | asked | Tariq Hussain | CC BY-SA 4.0 |