Closed ghost closed 3 years ago
Hi @Zaxorn , even though it can be hard, I have had better results on thin objects. Can you post the training_images (in a folder near your checkpoint)? It can happen that thin objects are not reconstructed well. In this case you can increase the BOOTSTRAP_RATIO parameter in the config to something like 16 and retrain
Here is the training images setting the BOOTSTRAP_RATIO to 16 (I don't know why the embedding is so wrong).
This is the model.ply I am using for training but it looks like fine, like the others I have already succesfully used cifarelli_1.zip
after how many iterations? I suppose this is at the very beginning.
You can also remove the occlusion augmentation to facilitate convergence. But it should work anyways.. What batch size did you use?
50000 iterations with a batch size of 8.
Here is the cfg file used for training. chiave_candela.zip
The picture is of the last iteration.
You need to set the BOOTSTRAP_RATIO to 16 and retrain the networks from scratch. If you only use batch size 8 you might need to decrease the learning rate as well.
Results look better on this object
but still wrong on the other one even if training looks good
Do I need to further increase the BOOTSTRAP_RATIO? Anyway what does this parameter do?
The bootstrap ratio refers to only calculating the reconstruction loss on the 1/bootstrap ratio pixels with the highest error. This forces the network not to just learn a constant background that is correct in most places.
Yes, you can further increase this ratio, but not too much. As I said the occlusion can hide important features in the input here so commenting out this line should help ' CoarseDropout( p=0.2, size_percent=0.05) ),'
Hello, I read from your paper that AAE fails for long and thin objects, but I have strange results also in these types of objects.
Is it possible to use AAE on these kind of objects or nothing? Thanks in advance