Feature Request: Output Management, Image Preview, and Face Swap Enhancements

[timfocus6]

First of all, thank you for all the hard work that has gone into this project. I truly appreciate the effort and dedication. I hope my suggestions don’t come across as intrusive, but I wanted to respectfully ask if there’s any chance we might see some improvements that could enhance the user experience. Here are a few ideas that I believe could be helpful:
1. Maintaining Output with Counting Numbers:
It would be great if there was an option to prevent the output from overwriting with each new run. Perhaps it could automatically append a counting number to the file names, so the previous outputs are preserved while the program is running.
2. Direct Saving of Preview Images:
I was wondering if it would be possible to have preview images saved directly, without needing any additional steps. This would definitely streamline the process, especially for quicker iteration and testing.
3. Face Resolution Matching:
If it’s possible, could we have an option to adjust the face resolution to better match the target image’s resolution? A feature that resizes the swapped face proportionally would be really helpful in maintaining the quality of the final image.
4. Improved Face Swapping for Hair and Customization:
One challenge I’ve noticed is that the face swap sometimes disturbs the target image’s hair, which can be quite noticeable. Would it be feasible to add a feature that lets users specify custom checkpoints or facial areas (such as the brow region) to make the swap more natural? It could help maintain a better blend, especially around sensitive areas like hair.

Again, I completely understand if these ideas may not be possible right now, but I thought I would respectfully bring them up for discussion. I’d love to hear any feedback from the community, and I look forward to seeing how the project evolves!

Thanks again for your incredible work!

[hacksider]

Hi there! Thanks for the feedback!

Let me do some quotations here to answer each suggestion properly.

1. Maintaining Output with Counting Numbers: It would be great if there was an option to prevent the output from overwriting with each new run. Perhaps it could automatically append a counting number to the file names, so the previous outputs are preserved while the program is running.

• This won't be a priority but I will add this on the list of improvements

2. Direct Saving of Preview Images: I was wondering if it would be possible to have preview images saved directly, without needing any additional steps. This would definitely streamline the process, especially for quicker iteration and testing.

• You can actually do a preview using live feature + OBS

3. Face Resolution Matching: If it’s possible, could we have an option to adjust the face resolution to better match the target image’s resolution?

• Will come out soon! We just need some time on this (but no option to adjust it manually or according to the resolution). It will just be better face than what we currently have now

4. Improved Face Swapping for Hair and Customization: One challenge I’ve noticed is that the face swap sometimes disturbs the target image’s hair, which can be quite noticeable. Would it be feasible to add a feature that lets users specify custom checkpoints or facial areas (such as the brow region) to make the swap more natural?

• Some will come some may not. We are focused on face as of this moment. So it's face centric for the next few months. Hair would be our last priority. Unless someone push some codes for this, then it will not be included on our releases (hoping some developers would help on this feature so we can focus more on our priorities as well).

[timfocus6]

Thank you so much for taking the time to respond to my suggestions! I really appreciate the clarity and the effort you’ve put into addressing each point. it means a lot.