search

dvirginz / DPC

37 stars 9 forks source link

I couldn‘t repreduce the result #2

Closed black-prince222 closed 2 years ago

black-prince222 commented 2 years ago

Hello，I try to use the checkpoints you provided to repreducing the results provided in the paper, but I can't not repreduce the result。 Here is the command I wrote: python train_point_corr.py --do_train False --gpus 0 --num_data_workers 8 --resume_from_checkpoint ./data/ckpts/shrec_ckpt.ckpt --default_root_dir ./output/

According to the paper, I should get 15.3% acc at 1% tolerance. But I only get 14.3%.

Could you tell me why? Thanks

itailang commented 2 years ago

Dear @black-prince222,

Thank you for your interest in our work. The accuracy may somewhat change due to the randomness of sampling points from the SHREC shapes.

black-prince222 commented 2 years ago

I think the code “pytorch_lightning.seed_everything(seed=seed)“ is supposed to ensure the sampling points are the same every time. Right? Actually, I got the same accuracy no matter how many times I test on shrec_ckpt.ckpt