Some questions about the code
xbq1994 opened this issue · comments
Boqiang Xu commented
Hi, I have some questions:
- As there are no descriptions about "task_type" in the annotations, how to split the 'data_part1_all_objaverse.json' and 'data_part2_scene.json' into train/val/test sets in proportion of (8:1:1)?
- I found you have uploaded files "voxelized_features_sam_nonzero_preprocess.zip" and "voxelized_voxels_sam_nonzero_preprocess.zip", what's that for?
- During the evaluation, the code needs the "coco_fmt_qust_file" annotations for computing accuracy. However there are no descriptions about "coco_fmt_qust_file" in the annotations, so how to calculate the test accuracy?
if hasattr(dataset[split], "coco_fmt_qust_file") and dataset[split].coco_fmt_qust_file is not None: self.ques_files[split] = dataset[split].coco_fmt_qust_file self.anno_files[split] = dataset[split].coco_fmt_anno_file
Yining Hong commented
- evaluation set: https://drive.google.com/drive/folders/1tdvEW5iB-ZEPo1tCqEAqKKhgHiDd4Dx0?usp=drive_link
- scannet features and point clouds
- we do not need that in the evaluation. We have uploaded our own evaluation scripts