Open BingoNate opened 8 months ago
I have the same requirement for this, and I hope the author can publicly disclose the training code. Thanks.
@BingoNate @se16wangmeiwei Dear friends, I realy enjoy EfficientSAM, so I try to implement the SAMI pre-training framework based on my understanding of EfficientSAM's SAMI framework and the technical details given in the paper. Unfortunately, I do not have the resources to test whether my implementation can reproduce the results reported in the paper. If you are still intersted at SAMI, you could try my codes.
My implementation of SAMI: https://github.com/yjh0410/SAMI
@BingoNate @se16wangmeiwei @yjh0410 Thanks for your interest in EfficientSAM. SAMI is implemented based on internal infra, and it may take a while to remove all the internal dependency. The implementation of SAMI is very close to MAE. When the code is ready, we will share it here.
@BingoNate @se16wangmeiwei @yjh0410 Thanks for your interest in EfficientSAM. SAMI is implemented based on internal infra, and it may take a while to remove all the internal dependency. The implementation of SAMI is very close to MAE. When the code is ready, we will share it here.
Amazing work! Hope the pre-training code can be released soon!
There are no training codes with SAMI Module since it's necessary for training custom datasets. Finetuning config settings and codes are also necessary for reproduce the results. Thanks.