Yoonho Na
Yoonho Na
@KevinXu02 Thanks for the comment. I just used rembg to remove the backgrounds in the images and trained nerfacto. but I found out that the results were even worse than...
@arbab-ml not yet. I think we should use mask when rendering somehow. I didn't had time to look up the code yet.
@arbab-ml thank you for sharing!
same
I haven't tried yet but you should check out the taming transformers repo there's a description about training in custom dataset. I think the overall framework is similar so it...
@Suimingzhe I experienced many trial and error for training ldm with custom dataset and successfully trained LDMs now and Yes the steps are correct. and in my opinion for the...
@Suimingzhe 1. At first, simply just compare reconstructed images with the actual input visually. If the reconstructed image show pretty much difference(It means that the pre-trained AE is not able...
@DongyangHuLi resolution was 1x256x256. I don't remember the exact time consumed for training each AE and LDM. training time depends on the dataset, batch_size, learning rate, ... and even hardware....
@zhangdan8962 Yes I did and I didn't observe big difference. Did you modified the training step function? If you did you have to modify the validation step in the same...