Closed hylin-work closed 3 years ago
The initial result looks very promising.
We would like to implement transfer learning using the existing DenoiSeg model and additional small size of site-specific data. Have you tried similar tasks or any example to share with us?
Thanks.
Can you please elaborate more on the transfer learning aspect? Do you mean training first with lots of noisy data for denoising and fine-tuning the trained network for segmentation with limited segmentation annotated data?
The initial result looks very promising.
We would like to implement transfer learning using the existing DenoiSeg model and additional small size of site-specific data. Have you tried similar tasks or any example to share with us?
Thanks.