This looks like a really cool (but also complicated) project.
@WhatsAI3 жыл бұрын
Indeed! Highly technical but so interesting!
@giallo_o3 жыл бұрын
this is one of the most groundbreaking things I have seen recently
@WhatsAI3 жыл бұрын
Indeed! It’s refreshing to see such a cool application! Text to image is such a fantastic task to tackle!
@giallo_o3 жыл бұрын
@@WhatsAI I make Plugins for Adobe Photoshop that applies certain effects on images, but that's like the "toddler" version compared to your samples
@pathanrageeb2 жыл бұрын
How to do it
@LukeSchoen3 жыл бұрын
Glorious!
@EskoLuontola3 жыл бұрын
Could this be used for creating better 180 videos for VR, so that you could move your head to change the perspective? Perhaps even walk around inside the scene. That would remove the biggest limitation of current 180 videos - a fixed viewpoint. For example have a grid of a dozen synced video cameras, all filming the scene from different perspectives, and then use ADOP to create a 3D model for each frame, and then play the 3D frames as 3D video or as an animated 3D model.
@WhatsAI3 жыл бұрын
I believe this could be an amazing solution for quick and effective 180 videos for Vr! It wouldn’t be perfect but it could give an amazing first “model” to work on for the animators/modelists/people that produce the videos and even be used alone sometime soon I am sure!
@tlatosmd3 жыл бұрын
Using this for interactive VR walk-throughs (interactive as such that you can freely move around it) was my immediate first thought as to what this entire project was all about.
@goldnutter4122 жыл бұрын
@@WhatsAI turns anything into metaverse ready Unity assets ? ;) cheers, great video
@MSI2k3 жыл бұрын
Holy moly! That is amazing
@WhatsAI3 жыл бұрын
Agreed haha! Couldn’t wait to share it!
@Morenob13 жыл бұрын
This gaming generation is gonna be amazing, I'm 34 now no way i'm gonna quit gaming. I thought about it until I bought a PS5
@WhatsAI3 жыл бұрын
Why would you quit if you love it! 😁
@Morenob13 жыл бұрын
@@WhatsAI I'm easily bored of games, it's a bit like when you are a kid playing with action figures but stops doing it once you're a teenager. But this gen doesn't make me wanna stop now. I've played Guardians of the Galaxy best looking game ever!
@vodkaru59073 жыл бұрын
Have to wait for them to publish the code
@WhatsAI3 жыл бұрын
Yep! I can update and ping you here whenever it is updated if you’d like! I’ll try to follow it closely :)
@trollenz3 жыл бұрын
Interested as well 👍🏻
@vodkaru59073 жыл бұрын
@@WhatsAI It's fine. I've bookmarked the github page to be on the look out. I have another question though. Do you know what software does the united spot channel use?
@luckys5413 жыл бұрын
@@vodkaru5907 Looks like the code has been released.
@WhatsAI3 жыл бұрын
You’re right, fantastic thank tou for pinging here!
@sinanrobillard28193 жыл бұрын
This is crazy! Possible to use for photogrammetry?
@Jojopiez3 жыл бұрын
That is so cool.
@WhatsAI3 жыл бұрын
I agree! I couldn’t wait to do a video about this research 😁
@jasoncysiu2 жыл бұрын
how much photos do you need to make them like a video, after training?
@XX-vu5jo3 жыл бұрын
I remember one of my senior researcher proposing something like this but he can't even code properly. He will just ask us to do it for him. LOL
@M_Jema47033 жыл бұрын
Amazing😍
@WhatsAI3 жыл бұрын
Agreed! 😁
@gandhamteja99183 жыл бұрын
I need a ai project for my final year.can u help in selecting one for me?
@WhatsAI3 жыл бұрын
Of course! Do you have any subfield or subject in mind?
@gandhamteja99183 жыл бұрын
@@WhatsAI Ai and ML
@WhatsAI3 жыл бұрын
Oh so nothing more precise haha.. what is your level of knowledge? What have you done yet related to these?
@gandhamteja99183 жыл бұрын
@@WhatsAI To be frank I did not have any knowledge related to it.I mean I didn't tried to learn it.But my team leader has some knowledge about it.can you suggest a medium level project
@WhatsAI3 жыл бұрын
Hmm sure, I’d say an image classification task would be the most appropriate for this case! It can be on any type of dataset available online and you can implement different architectures and see what works best in your specific task. You can find many datasets or similar challenges on kaggle or online elsewhere and it can be a great final year project if you attack an interesting task like trying to detect tumors in medical images or useful industry detection/classification like defect pieces or not