search

drprojects / superpoint_transformer

Official PyTorch implementation of Superpoint Transformer introduced in [ICCV'23] "Efficient 3D Semantic Segmentation with Superpoint Transformer" and SuperCluster introduced in [3DV'24 Oral] "Scalable 3D Panoptic Segmentation As Superpoint Graph Clustering"
MIT License
508 stars 65 forks source link

stuff_classes #125

Closed ImaneTopo closed 2 weeks ago

ImaneTopo commented 2 weeks ago

hello Mr.Damien, during the training of the model on my own dataset whish contains 2 classes : pomme(apple) and clutter, normally the instances are apples, I don't know how to fix the stuff_classes image

drprojects commented 2 weeks ago

If you are doing apple panoptic segmentation, you should have set your dataset to have 'clutter' as the only 'stuff' class and apple is, as a result, the only 'thing' class.

Have a look at the code for BaseDataset.stuff_classes, it is fairly documented and straightforward. See also the provided datasets and how 'thing' and 'stuff' classes are declared and handled.

If you ❤️ or use this project, don't forget to give it a ⭐, it means a lot to us !