xmed-lab / CLIPN

ICCV 2023: CLIPN for Zero-Shot OOD Detection: Teaching CLIP to Say No
MIT License
116 stars 10 forks source link

Could you release the code for CIFAR100 dataset? #4

Closed cailincan0129 closed 7 months ago

cailincan0129 commented 11 months ago

Really appreciate your work. It is really impressive. I'd like to replicate the results using CIFAR-100 as the in-distribution dataset. Is it possible to access the code for the CIFAR-100 dataset or receive guidance on image transformations and dataset splitting, such as using CIFAR-10 as the out-of-distribution dataset? Thank you very much.

SiLangWHL commented 11 months ago

Hi,

Thanks for your interest in our work. All experiments on Cifar are conducted based on CLIPN with hand-crafted prompts. The original plan is to release those codes this month. However, I found it hard to free up my time to reorganize them until mid-November due to several very important things, such as preparing for my PhD PQE. I will release them before the end of November. When the codes are released, I will contact you!

Bests, Hualiang

cailincan0129 commented 11 months ago

OK! Thank you very much!

saya0312 commented 9 months ago

Sorry to bother you ... I am interested in CLIPN and am eagerly awaiting the release of the code using CIFAR-10 and CIFAR-100. When will it be released?

SiLangWHL commented 9 months ago

Hi, the code is coming. I am writing some necessary code comments. I will upload it tomorrow.

saya0312 commented 9 months ago

Thank you very much!

SiLangWHL commented 9 months ago

Hi, codes are released. Although there are still some hard codes that need to be optimized further, the main body of the code is ready to run.