Pier Carlo Cadoppi
Pier Carlo Cadoppi
Hi, did you try it out on the public version? (The one available on overleaf.com)
That sounds perfect, thanks a lot!
I have this issue too!
> I refined the bboxes' quality using the segmentation maps. > > https://github.com/hrsma2i/modanet#Setup hi, can you show us an example of a refinement you made?
hi guys, I have a much better version of the annotations json wich you can download [here](https://drive.google.com/open?id=1CVWJ-QYhXV6y58gsV3XBu2gnfjfTKjcz) it tries to merge and move the annotations of footwear and boots, by...
By that you mean that I copied some annotations from one image to another one inadvertently? You can reconstruct the file by using the original one and applying the script...
Hi guys! @eliorc @rakhecha42 @Zahra-Saki @change2014 I coded the solution in a nice package that installs automatically and allows to train this dataset using [Mask R-CNN](https://github.com/fizyr/keras-maskrcnn). Below is the one-liner...
@simaiden I too had the same problem two days ago. Since I have the original files on my desktop computer I'll upload them to my repo and let you know!
https://github.com/cad0p/maskrcnn-modanet/releases/tag/v1.0.3 Done, please consider trying my repo too! @simaiden
Only 50k are annotated :) On Fri, Oct 18, 2019, 03:19 Amal wrote: > Thanks for facilitating downloading the dataset, did you annotated the 1 > million images or only...