Looks great! I roughly tried v1.6.2 and I noticed it's not accepting .AVI video containers (while on mp4 H.264 for example) it just being stuck on "Loading Video... " Will you add more video formats and containers in the next update? Sp,e suggestions: 1. At the moment it doesn't show the progress so it's hard to tell how long it will take to finish, Estimate Time + Percentage will be nice to have while processing ANY of the AI options. At the moment it's in the background, the current panel won't show it, but it looks nice on v2 as you show in this video 👍 2. When loading a VIDEO file, can you add autodetect FPS so the user won't need to manually type the FPS of the video? Thanks ahead, and keep up the good work!
@wladradchenko
5 ай бұрын
Thanks for sharing your thoughts and observations! Regarding .AVI format support, I know about this problem in old version, but I don't know the new version will work with .AVI files or not. I need to check this if this format is important. I've focused on optimizing compatibility with .mov and .mp4 formats, but I'm always looking to expand the supported formats based on user feedback. I've made improvements to FPS detection in the latest version, fixing some previous issues. As for displaying the processing time until completion, it's a bit tricky to implement accurately, but I'll explore options to make this feature more user-friendly in future updates. In the new version, I've changed how progress is shown to prioritize what's actively running and being processed and by feedback I will improve show progress on other versions. Your feedback is valuable, and I appreciate your support for the project!
@MrDanINSANE
5 ай бұрын
@@wladradchenko Thank you for your detailed reply! I sure have more ideas but I will wait for v2 since of many changes you already mentioned and shown in the video then I will share some of my suggestions to help you improve it if it will be possible of course. Keep up the amazing work you do! 💙
@mybiggestdreamsfulfilled1028
4 ай бұрын
Add webcam to change skin to on live video call not just face😊
@WaitingYi-ye8qr
5 ай бұрын
will v2 open source too?
@wladradchenko
5 ай бұрын
Probably yes. It all depends on how it will be received and whether users are waiting for it. But I think the way out will be tamed to some event. For example, 1000 stars on Github or some kind of holiday. I didn't published updates yet. I need to test on difference OS, how will be work and do small fix. For example , yesterday I found out what custom fonts not show correctly on Windows and need to change on default, I don't know why yet.
@zeropoint1776
5 ай бұрын
thank you, can you create IOS app for this?
@wladradchenko
5 ай бұрын
I think is not necessary because this has mobile pages, you will able to use IOS browser to use Wunjo by public link
@eccentricballad9039
5 ай бұрын
I just downloaded V1. I love it so so much. People don't know how much time and effort it must have took you to make this easily available to everyone for free. Thanks a lot, I really really appreciate you. You are a godsend, an angel. PS : I'm looking forward to V2
@wladradchenko
5 ай бұрын
Thank u so much that you use app and waiting for new version ❤
@eccentricballad9039
5 ай бұрын
@@wladradchenko (gonna delete this comment later) I have an unusual problem. i started faceswap on a 10 second video which took me around 15 minutes on my GPU 1650ti 4GB VRAM. I was so curious that i started checking folder as the images loaded one by one, frame by frame to around 326 frames after that all the process went on and face was swapped frame by frame and then converted into the .mp4 file in around 15 minutes but i saw that i didn't get any message/ notification in logs and it went one. So i copied the video which was created on 15 minute mark, which was perfect the face was swapped perfectly, frames were perfect and the resolution too but then the logs kept on going and going for another 30 minutes and then i got this at the end Face swap synthesis completed successfully! Unable to determine image format: cannot identify image file 'C:\\Users\\Meena\\.wunjo\\tmp\\video_1714824391581_Xrkcj_cut.mp4' 100%|##########| 326/326 [30:00
@tribuzeus
5 ай бұрын
Hi, Very Nice! But clone Voice in portuguese Brazilian too? Thanks
@wladradchenko
5 ай бұрын
In the new release, I have not yet added voice cloning, since the previous approach needs to be changed completely, it is complicated and takes a whole update. If this update is accepted for a long time, I will start adding features for working with sound, including improving the audio separator, and adding new languages, for example, to the interface and to the voice clone, but now the solution will be universal and have many languages
@tribuzeus
5 ай бұрын
@@wladradchenko ok, and what are the minimum hardware requirements?
@wladradchenko
5 ай бұрын
@@tribuzeus if use only CPU than min 8 Gb available RAM, and 10 Gb disk space for app and 10-40 Gb for models. Neural models will be downloaded by first time run of module. If use GPu, than minimum available 8Gb VRAM and 8-12 RAM. In settings user can see available RAM and VRAM.
@reflectionsAND
5 ай бұрын
Thanks, so detailed and entertaining instructions too! 🎉
@Vishwajeet987
4 ай бұрын
Very Excited! Any update on the date this version would launch?
@wladradchenko
4 ай бұрын
I think 24 May on my birthday)) In fact, everything is ready for launch, except for this github.com/beeware/briefcase/issues/1788. I would like to make an installer for the GPU version, and not an archive as before.
Пікірлер: 20