Open AIMads opened 2 years ago
Yes, if you want to train the network on a new dataset, you can just reimplement the function to read the camera intrinsic. https://github.com/EryiXie/PlaneRecNet/blob/cea1e8b1edf054f59e15891fa799cdcc5feda72d/data/datasets.py#L174
But for inferencing with any data samples, the camera intrinsic is not asked to be given. However, rumors said neural networks learn the camera parameters implicitly if trained for monocular depth estimation in the most vanilla ways, and therefore they can not generalize well for other camera models.
Hey!
Is it possible to change intrinsic parameters to fit with your own camera? Or is the model not affected by it?