search

charlesq34 / frustum-pointnets

Frustum PointNets for 3D Object Detection from RGB-D Data
Apache License 2.0
1.59k stars 538 forks source link

matlab interference for SUNRGBD dataset #34

Open shuluoshu opened 6 years ago

shuluoshu commented 6 years ago

Hi, @charlesq34 I find a problem in using the matlab code extract_rgbd_data.m under SUNRGBD toolbox, since there is no way to deal with the training and testing separately. In other words, I cannot get the testing dataset and the val_data_idx.txt by your offered matlab code. Besides, I can't find the folders SUNRGBDtoolbox/traintestsplit , how do you generate it? thanks a lot!

wangszoo commented 6 years ago

Hi， @shuluoshu I have a same question ,have you solved this problem? As your question ,can I think val_data_idx.txt is generated from extract_rgbd_data.m? But run extract_rgbd_data.m, I only get four floder: depth, image, calib and label. So I want to know how did you solve it？At the same time , I find trainvalspilt in SUNRGBDtoolbox/traintestRGBD/allsplit.mat,Is this helpful for solving this problem？ thanks a lot!

PranjalBiswas commented 6 years ago

Hi @charlesq34 , @shuluoshu , @wangchau

I was trying to generate the pickle files for SUNRGBD, but running ¨sunrgbd_data/sunrgbd_data.py¨ gives an error. If I understand correctly I need to have train/val file list. Cause after running ¨extract_rgbd_data.m¨ I just get 4 empty folders. Can someone guide me what needs to be done to get the pickle files. Thanks a lot in advance.

Kind Regards Pranjal Biswas

Update: This should help: https://github.com/charlesq34/frustum-pointnets/issues/51#issuecomment-426581516

ybyangjing commented 5 years ago

hi @charlesq34 , @shuluoshu , @wangchau @PranjaLBiswas27 I use extract_rgdb_data.m to generate 4 folders and it not empty. but when I was trying to generate the pickle files to train SUNRGBD. I run sunrgbd_data/sunrgbd_data.py and given an error. I have no train_data_idx.txt files.Could you guide me what need to be done to get the pickle files. Thanks a lot in advance. Regards, Jing Yang

shengrongjin commented 5 years ago

@shuluoshu @wangchau @PranjaLBiswas27 @ybyangjing @cclauss Have you solved the problem? Can you tell me how to solve it? Thank you very much !

PranjalBiswas commented 5 years ago

Hi everyone, please have a look at the update on my previous comment in this thread. I hope that helps. Success!!!

shengrongjin commented 5 years ago

@PranjaLBiswas27 My friend, thank you for your reply very much. I'll go and have a look at it now.