search

advimman / lama

🦙 LaMa Image Inpainting, Resolution-robust Large Mask Inpainting with Fourier Convolutions, WACV 2022
https://advimman.github.io/lama-project/
Apache License 2.0
7.91k stars 838 forks source link

how to make my own dataset #294

Open yangyuke001 opened 8 months ago

yangyuke001 commented 8 months ago

Hello, thank you for your excellent work！ May I ask if I want to train a model specifically for removing watermarks? The background images are all product images from e-commerce platforms. My current idea is as follows: I plan to collect a certain amount of e-commerce image dataset, each image without watermarks, randomly generate masks using the author's methods, and further create a training set. Please help me check if my idea is correct?

LikeGiver commented 2 months ago

I'm also interested in this idea, my current model is based on this project: https://github.com/lcy0604/EraseNet , but the effect is not satisfactory. I think it is due to the small number of model parameters.