search

PRBonn / OverlapNet

OverlapNet - Loop Closing for 3D LiDAR-based SLAM (chen2020rss)
MIT License
653 stars 115 forks source link

Pytorch version training. #33

Open lymdove opened 2 years ago

lymdove commented 2 years ago

Thanks for your great work! I want to compare my lopp closure method with Overlap, but I have problem with running the work. I use the pytorch version, and I already generate depth, intensity, normal data. I want to compute a score for every pair, but I have no model weights. So I want to train one, but I can not find which file could generate the 'overlaps/train_set.npz'. Could you help me, or could you provide a pre-model. Thanks.

Chen-Xieyuanli commented 1 year ago

@BIT-MJY could you please provide the pre-trained model using PyTorch?

@lymdove you may also have a look at our novel overlaptransformer: https://github.com/haomo-ai/OverlapTransformer It is implemented using PyTorch, and the pre-trained model was given.

surajiitd commented 1 year ago

@BIT-MJY could you please provide the pre-trained model using PyTorch?

surajiitd commented 1 year ago

@lymdove demo/demo4_gen_gt_files.py can generate the train_set.npz file in data/{seq_num}/ground_truth/train_set.npz

laebe commented 2 months ago

I'm not sure, because I was not involved in the pytorch version. But my guess is that the pytorch version uses the input which can be generated by the non-pytorch version (as stated above e.g. with demo4)