search

yuanze-lin / Learnable_Regions

[CVPR 2024] Official code for "Text-Driven Image Editing via Learnable Regions"
https://yuanze-lin.me/LearnableRegions_page/
MIT License
266 stars 21 forks source link

Code release? #1

Closed Decem-Y closed 8 months ago

Decem-Y commented 8 months ago

good work！！！ your project looks very interesting, do you have a date for it's release?

yuanze-lin commented 8 months ago

good work！！！ your project looks very interesting, do you have a date for it's release?

Hello, thank you for your attention, I'm now working on other projects and catching up with incoming conferences, I'll release the code soon

Decem-Y commented 8 months ago

good work！！！ your project looks very interesting, do you have a date for it's release?

Hello, thank you for your attention, I'm now working on other projects and catching up with incoming conferences, I'll release the code soon

Thanks for the reply, I have one more question, is there a suggested training dataset to use?

yuanze-lin commented 8 months ago

Thanks for the reply, I have one more question, is there a suggested training dataset to use?

You're welcome, actually, I think there are some options, it depends on you:

(1) You can collect the images from a free image website, like unsplash. (2) You can use the TEdBench. (3) You can also adopt the dataset from instruct-pix2pix.

If you want to evaluate your method, I think you can use TEdBench or the collected images from unsplash.

Decem-Y commented 8 months ago

Thanks for the reply, I have one more question, is there a suggested training dataset to use?

You're welcome, actually, I think there are some options, it depends on you:

(1) You can collect the images from a free image website, like unsplash. (2) You can use the TEdBench. (3) You can also adopt the dataset from instruct-pix2pix.

If you want to evaluate your method, I think you can use TEdBench or the collected images from unsplash.

Thank you for your guidance, it was very helpful for me. I wish you success in your incoming conferences.

yuanze-lin commented 3 months ago

Thanks for the reply, I have one more question, is there a suggested training dataset to use?

You're welcome, actually, I think there are some options, it depends on you: (1) You can collect the images from a free image website, like unsplash. (2) You can use the TEdBench. (3) You can also adopt the dataset from instruct-pix2pix. If you want to evaluate your method, I think you can use TEdBench or the collected images from unsplash.

Thank you for your guidance, it was very helpful for me. I wish you success in your incoming conferences.

Very sorry for the delay in releasing the code. The code has been released now, so have fun playing with it! :)