Closed af-doom closed 6 months ago
Our implementation was validated in only the Ubuntu environment. As a guideline to use the pretrained models, we will commit example codes for stitching of one's images maybe til this month.
Sorry for the delay. Now we have added example codes.
Thanks.
Hello, thank you very much for your great open-source work. Does the current code only support ubuntu? How to use a pre trained model to concatenate one's own images