ValueError: could not convert string to float: '-9.1\x160'

[whuhxb]

When I running the code: python partition/partition.py --dataset s3dis --ROOT_PATH $S3DIR_DIR --voxel_width 0.03 --reg_strength 0.03 I encounter this bug on hallway2 of Area3: xyz = np.ascontiguousarray(room_ver[:, 0:3], dtype='float32') lib/python3.6/site-packages/numpy/core/_asarray.py", line 177, in ascontiguousarray return array(a, dtype, copy=False, order='C', ndmin=1) ValueError: could not convert string to float: '-9.1\x160'

Has anyone encountered the same problem?

[loicland]

Hi,

there are errors in the S3DIS dataset. We provide a patch to remove some of these errors, see:
https://github.com/loicland/superpoint_graph/blob/ssp%2Bspg/S3DIS.md

Let me know if the problem persists.

[loicland]

Hi!

We are releasing a new version of SuperPoint Graph called SuperPoint Transformer (SPT). It is better in any way:

✨ SPT in numbers ✨
📊 SOTA results: 76.0 mIoU S3DIS 6-Fold, 63.5 mIoU on KITTI-360 Val, 79.6 mIoU on DALES
🦋 212k parameters only!
⚡ Trains on S3DIS in 3h on 1 GPU
⚡ Preprocessing is x7 faster than SPG!
🚀 Easy install (no more boost!)

If you are interested in lightweight, high-performance 3D deep learning, you should check it out. In the meantime, we will finally retire SPG and stop maintaining this repo.