Hi, I am trying to add several pre-processing steps to the ComputerVision onnx I am working on. I was able to add almost all the pre and post-processing steps I wanted, but now I want to solve the case where the input image has an alpha channel.
So the original model expects [ batch, 3, width, height ], but the application I am working on deals with images that have an alpha channel [ batch, 4, width, height ].
How can I go about adding a pre-processing step inside the onnx model that would be able to drop the alpha channel?
Description
Hi, I am trying to add several pre-processing steps to the ComputerVision
onnx
I am working on. I was able to add almost all the pre and post-processing steps I wanted, but now I want to solve the case where the input image has an alpha channel.So the original model expects
[ batch, 3, width, height ]
, but the application I am working on deals with images that have an alpha channel[ batch, 4, width, height ]
.How can I go about adding a pre-processing step inside the
onnx
model that would be able to drop the alpha channel?Thank you for your help :)