I have an ONNX model that takes an image as input, transforms it using a computational model, and outputs a transformed image. I can explain more what the model does if anyone wants to know more!
But TLDR I want to get the camera texture from the VR pass through cameras to be used as an input and then display the output of the model on the headset.
I was using Barracuda until now and it was working pretty well but soon I hit a wall with more complex ONNX models. now I can’t even run my old onnx models with sentis.
Anyone working on this kind of project? would love to get some insights, specifically how to convert an image output tensor to a texture
2 posts - 2 participants