search

Ivan-Tang-3D / Point-PEFT

(AAAI2024) Point-PEFT: Parameter-Efficient Fine-Tuning for 3D Pre-trained Models
41 stars 2 forks source link

Code about Point-MAE #2

Open ZewLin opened 2 months ago

ZewLin commented 2 months ago

Hello dear authors, thanks for your amazing work! I find the current code use Point-M2AE as backbone, can you release the code use Point-MAE as backbone? Thanks!!!

Ivan-Tang-3D commented 2 months ago

Hello dear authors, thanks for your amazing work! I find the current code use Point-M2AE as backbone, can you release the code use Point-MAE as backbone? Thanks!!!

Thanks for your appreciation. Since I am busy these few days, I will release the related code for Point-MAE in the near future. Once it is released, I will notify you.

ZewLin commented 2 months ago

Hello dear authors, thanks for your amazing work! I find the current code use Point-M2AE as backbone, can you release the code use Point-MAE as backbone? Thanks!!!

Thanks for your appreciation. Since I am busy these few days, I will release the related code for Point-MAE in the near future. Once it is released, I will notify you.

Thanks for your reply!!!

ZewLin commented 2 months ago

I have another question, did you reproduce the full-parameter finetuning in modelnet and scan dataset based on Point-MAE and reach to the result as mentioned in paper? Looking forward to your reply! Thanks again!

Ivan-Tang-3D commented 2 months ago

I have another question, did you reproduce the full-parameter finetuning in modelnet and scan dataset based on Point-MAE and reach to the result as mentioned in paper? Looking forward to your reply! Thanks again!

Yes, the results I reproduced are close to the performance as its paper mentioned.

Ivan-Tang-3D commented 2 months ago

I have another question, did you reproduce the full-parameter finetuning in modelnet and scan dataset based on Point-MAE and reach to the result as mentioned in paper? Looking forward to your reply! Thanks again!

Yes, the results I reproduced are close to the performance as its paper mentioned.

OK,thanks! I am also research in point pre-training, can we add a contact information to communicate in detail? My wechat account is wxid_swxw9jnfctiu22. Thanks again for your reply and help!!!

Sorry, the wechat account displays unavailable.

ZewLin commented 2 months ago

I have another question, did you reproduce the full-parameter finetuning in modelnet and scan dataset based on Point-MAE and reach to the result as mentioned in paper? Looking forward to your reply! Thanks again!

Yes, the results I reproduced are close to the performance as its paper mentioned.

OK,thanks! I am also research in point pre-training, can we add a contact information to communicate in detail? My wechat account is wxid_swxw9jnfctiu22. Thanks again for your reply and help!!!

Sorry, the wechat account displays unavailable.

Oh sorry , it's my mistake. Could your retry the account 19550151227? Thanks!

ZewLin commented 1 month ago

I have another question, did you reproduce the full-parameter finetuning in modelnet and scan dataset based on Point-MAE and reach to the result as mentioned in paper? Looking forward to your reply! Thanks again!

Yes, the results I reproduced are close to the performance as its paper mentioned.

OK,thanks! I am also research in point pre-training, can we add a contact information to communicate in detail? My wechat account is wxid_swxw9jnfctiu22. Thanks again for your reply and help!!!

Sorry, the wechat account displays unavailable.

Oh sorry , it's my mistake. Could your retry the account 19550151227? Thanks!

@Ivan-Tang-3D Hello, I'm sorry to bother you. Is this account available?