Anh Nguyen

Hi, it seems something missing in your dataset. Please check [Issues 8](https://github.com/nqanh/affordance-net/issues/8) for a recent discussion.

No, it's not correct. For each object in the xml file, you need a mask groundtruth. > In general, each object in the .xml file must have a groundtruth mask...

Yes, the number of .sm files must equal to the number of objects in the xml file.

Yes, pretrained model helps us to reduce the training time and also better initialization for the model parameters. If you don't use it, then I guess you will need to...

Hi, AffordanceNet code includes some custom layers for Caffe. If you use the original Caffe that comes with Docker, you don't have these layers. Therefore, it would be more convenient...

You can find more information here: https://github.com/nqanh/affordance-net/issues/8

Please see https://github.com/nqanh/affordance-net/issues/26

First, just create a transparent image using OpenCV of the final affordance map over the original image. Then, you can copy and replace the object in the original image with...

For the camera parameters, it depends on where the images are collected. The images are collected from three kinds of camera: - Asus xtion camera - MultiSense SL - From...

Do you train on Pascal VOC dataset (not IIT-AFF)? If you train on a different dataset, then you have to prepare the input in the right way. Please see Readme...