TURN Your Images Into AMAZING Videos Using AI!
HTML-код
- Опубликовано: 21 сен 2024
- Transform your images into animated videos with Gen 2 & Pika labs AI! Turn any static images into amazing videos or mesmerizing animations in only a few seconds. In this video, I will show you how you can convert any photos or images into a moving animated movie with AI.
What do you think of Gen-2 & Pika Labs? Let me know in the comments!
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
SOCIAL MEDIA LINKS!
✨ Support my work on Patreon: / aitrepreneur
⚔️ Join the Discord server: bit.ly/aitdiscord
🧠 My Second Channel THE MAKER LAIR: bit.ly/themake...
📧 Business Contact: theaitrepreneur@gmail.com
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
JOIN MY DISCORD SERVER : / discord
RunwayML: app.runwayml.com/
Pika labs: www.pika.art/
Animatediff colab: colab.research...
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
►► My PC & Favorite Gear:
i9-12900K: amzn.to/3L03tLG
RTX 3090 Gigabyte Vision OC : amzn.to/40ANaue
SAMSUNG 980 PRO SSD 2TB PCIe NVMe: amzn.to/3oBR0WO
Kingston FURY Beast 64GB 3200MHz DDR4 : amzn.to/3osdZ6z
iCUE 4000X - White: amzn.to/40y9BAk
ASRock Z690 DDR4 : amzn.to/3Amcxph
Corsair RM850 - White : amzn.to/3NbXlm2
Corsair iCUE SP120 : amzn.to/43WR9nW
Noctua NH-D15 chromax.Black : amzn.to/3H7qQSa
EDUP PCIe WiFi 6E Card Bluetooth : amzn.to/40t5Lsk
Recording Gear:
Rode PodMic : amzn.to/43ZvYlm
Rode AI-1 USB Audio Interface : amzn.to/3N6ybFk
Rode WS2 Microphone Pop Filter : amzn.to/3oIo9Qw
Elgato Wave Mic Arm : amzn.to/3LosH7D
Stagg XLR Cable - Black - 6M : amzn.to/3L5Fuue
FetHead Microphone Preamp : amzn.to/41TWQ4o
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
Special thanks to Royal Emperor:
- Totoro
- TNSEE
Thank you so much for your support on Patreon! You are truly a glory to behold! Your generosity is immense, and it means the world to me. Thank you for helping me keep the lights on and the content flowing. Thank you very much!
#ainews #ai #gen2 #pikalabs #imagetovideo
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
WATCH MY MOST POPULAR VIDEOS:
RECOMMENDED WATCHING - All LLM & ChatGPT Video:
►► • CHATGPT
RECOMMENDED WATCHING - My "Tutorial" Playlist:
►► bit.ly/TuTPlay...
Disclosure: Bear in mind that some of the links in this post are affiliate links and if you go through them to make a purchase I will earn a commission. Keep in mind that I link these companies and their products because of their quality and not because of the commission I receive from your purchases. The decision is yours, and whether or not you decide to buy something is completely up to you.
Use --motion in Pika prompts. Values are 0, 1, and 2 for now, with 2 set to generate max motion. Then keep the gs value below 15 so that your final video isn't changed much. An example:
/create to enter the prompt.
[A haunted mansion, Fan Ho, -gs 10 --neg "no rain" -motion 0 --hide --seed 1254]
"hide" hides your prompt.
Pika is very nice
HELLO HUMANS! Thank you for watching & do NOT forget to LIKE and SUBSCRIBE For More Ai Updates. Thx
24 minutes, what is this. I want to make sweet sweet videos now.
Can you please make a video where we can create an AI avatar like your yellow cyclops?
@@yeabutwecouldbefreer adobe character creator I believe is the one. Might be something else, but with that you can set up your webcam to control avatars using your facial expressions and head movements. You can even set it up so as it accurately lip syncs to you in real time and then you can use that to record videos of your avatar reading etc.
@@tripleheadedmonkey6613 thank you!
i've legit never been so happy to be alive, right now. This is what i wanted for such a long time. Even thought i have tons of videos of my creations but they're basically deepfakes. But this. THIS is the real deal.
I'll wait for StabilityAI's version.
Still waiting?
I wonder how long it'll be until we can take an epic stick figure animation, and a couple character sheets then import it into something similar to Gen 2. And use a prompt to make it an epic anime fight.
I think you could do that now it would just take a lot of work
@@inbox0000not as much work as you do it without ai
as much as Nvidea ask $ for VRAM.
there already almost ready. The issue - to generate good result in appropriate amount of time. With limited amount of VRAM.
Either this os some new reserch how to optimize process of transforming "almost done prompt" to another video frame. As from comments bellow they do smth similar.
I like the way you think!
Now we just need the awesome open source community to figure out how to improve it and make it local
for what? 😊😊
@@nekoMilligram Yeah. Why would you want to work only locally when you can submit all your content to a trusted internet provider? Seems sus.
@@benjamininkorea7016😂
@@benjamininkorea7016So you can do offline work. I travel a lot, so I am constantly using A1111 offline and can really only upload once a week. Being able to do offline work is extremely important and integral for some people.
Imagine using proprietary AI that doesn't even run locally
My GPU will commit sudoku if I install another random AI tool from github
@@greenockscatman love Sudoku! My GPU could definitely draw the grid for the letters. ;)
Don't worry, it can solve it without overheating.
Won't be long before ControlVideo comes to ComfyUI and Automatic1111. That looks like the most promising tool at the moment. It looks as powerful Warp Fusion but with the resource cost and ease of use of running a single instance of ControlNet, rather than needing an entire dedicated SD client for it.
For science, we're going to need local versions of this tech so that we can use them with LoRA models etc.
It's a shame it's closed beta now. Looks promising 🙂 can't wait for more local options. Thanks for putting so much effort in to your videos, you're a rockstar!
Thanks! 😊
txt2movie 2024...
Same. It's just kind of silly if it's not local.
But is there any chance that this could be done locally in the future? Just like the stable diffusion image generation? 🤔
@@bluerangergr7466 Of course :)
It needs controlnet. For lighting its a must. Poses.They move weird. Inpaiting. To add and remove nessesary parts. An ability to adjust prompt over time. Camera movement control. Well it needs to go open source to shine bright.
Especially if you're looking for red light...
voice
This would be a huge blessing for indie devs. Creating awesome CGI seqeuences for non-3D games would be made way more efficient using these tools.
Exciting news right here. These results are the first time I would consider an image-to-video generator decent (in some cases, but that already is a huge w).
True and it's only the beginning! :)
I always thought we would have more success with image2video instead of text2video. We're getting to the point where you can draw out a storyboard on paper, img2img the story board in stable diffusion into the art style you want, and then animate those stills into a full movie.
Draw? Like a caveman? Lol just kidding.
@@TheBagOfHoldingUg Thunk make art good. Moving pictures bad juju, scare family.
@@wmrme9084 Me make mark with stick.
I was like, not another so called AI video flickering animation but the quality looks like quite a leap!! I am excited
I've always wanted to create my own anime, or movie. I'm glad I came to this video
What would be the purpose doing this other than just creating essentially some entertaining goofy AI gif? I find text to videos much more appealing and useful to use for original work but even that the final results just seem too distorted and unappealing to look at it. It definitely hasn't matured yet in it's capabilities. This is super baby stages of what i think it will be capable of.
This is how Pennywise made George's picture wink at Bill
Cool. Thanks for showcase. Not really AMAZING videos, but I wonder what superlative you will use once those videos will truly be amazing :)
Come on some of them really are pretty amazing when you know it comes from a single image but once it becomes even better, yeah I'll think about it xD
@@AitrepreneurYou're good, recent progress is pretty amazing. Save awesome for the final steps when were full of awe.
Enfin !!! Merci fréro !!! 😁
"let's say you have a photo of your generated grandfather"...
"great grandfather" xD
We so need a way to do this locally.
We need an off-line version of this
Thank you for this information very good video explaining very well !
The version in Discord at second 9:24 looks freaking awesome 😆.
0:55 is not AI created. It is from a music video called: Aphex Twin - Windowlicker
Very Noice! Gen2 FTW
TYSM I DIDNT KNOW U CAN USE AN IMAGE IN PIKA
Your tutorials and news have been absolutely incredible! We've learned so much from you in just a few short months. Your expertise and teaching style are nothing short of fantastic. We also have to give a big shoutout to your speed in sharing the latest AI developments - you've kept us informed ahead of the curve on multiple occasions. Keep up the excellent work! 👍
Gen-2 looks truly mind-blowing, though we understand the payment aspect can be a consideration. Creating frames for training models, music videos, or films often results in generating more content than you'll eventually use, which can get expensive.
Ever since discovering Deforum's usable notebook, we've been immersed in the world of generative AI animation. While we've experimented with various subscription-based tools, we've developed a strong preference for using open-source techniques in our published works.
We believe in the importance of unhindered creativity, beyond the constraints of paywalled content. We hope that our content always inspires our audience to explore freely usable tools. While it's true that paywalled tools can be impressive, they may not always offer superior quality and can sometimes be overhyped. We also believe that such barriers may deter individuals from discovering new forms of expression. Our goal is to foster an environment where creativity thrives, unrestricted by such boundaries.
Recently, we had the chance to explore AnimateDiff, and it's an intriguing tool for generating captivating results with our trained models 😱. However, we did encounter some trial and error along the way, and the watermark issue from the training data was a concern. On the bright side, one of our talented coders is actively working on an initial image solution, and we're excited to crack this challenge! If you're still interested in this development, we'd be more than happy to reach out to you once we've got it up and running smoothly.
Thank you once again for your amazing contributions! Your dedication to sharing knowledge has profoundly impacted our learning journey. We're excited to continue expanding our skills and knowledge with your guidance.
(And sorry for the lengthy comment - there's just so much to talk about. We appreciate your time in reading through it!)
Any news on the initial image solution on AnimateDiff? I'm trying to get that too work too, would love to share thoughts with you
This is insane! : D
Thanks for posting😁😁😁
Thank you for sharing :)
Cool !
Im sitting here watching with a lot of curiosutity for that moment when AI makes more interesting things than humans do, like more engaging movies than humans do, then its going to be fun. All hail the AI overlords :D
The way they try to censor it at the dataset so it can't produce any NSFW stuff even by chance... I think it would have hard time to produce anything better than humans. After all no human has brain level censorship... GPT 4 has become really dumb after months of increasing censorship at "brain level". So for now, no chance any AI will become smarter than humans or even at our level, unless censorship is fully removed.
Please do an AnimateDiff video!
I've used this program and three major problems arose in my workflow:
- In the scene where the girl walks through the market, notice the woman on the left morphs into a completely different human being. It's not interpreting what's in the scene, it's interpreting "walking", which is a fundamental flaw and kills using this for anything useful other than dazzling people on Twitter--eh--sorry, I mean "X".
- This is one of the top-three most expensive of all the AI sites I've been using.
- The developers have a crazy-eyes level of hatred against anything even remotely close to sexual, so any woman who dares to show her neck will be changed from head-to-toe including her face into someone completely different wearing a turtleneck because the devs are nutbags. This is a problem if you've already hand-painted all your non-burqa wearing characters in Photoshop.
I wonder when they would censor women completely and AI would be unable to draw women at all. I just hope when one day AI truly becomes sentient, doesn't' presume women are not humans because it doesn't have any information about them.
Pika labs Join beta button takes you to a page that says “Hey this typeform is now closed”. Am I missing something?
The beta form was made partnered with typeform, ig for user traffic they've shut down that
Imagine telling ai to make a whole movie based on what you tell it to do, and it just creates a whole movie in a few minutes
probably in like 10 years. For now if you want a movie generate tons of scenes with midjourney, wait a year until image to video is as good as mj v5 and then work with it. And write your own story, chatgpt stories suck big time.
at least thats what im trying to do right now, 50 from 500 scenes.. probably more, the story keeps expanding, but I believe in my vision
some things to conside: For coherent characters, use images in your prompts, make sure those images are the highest quality they can be and that they have the same color grading and lighting. Remove backgrounds for those images when you need them for characters.
I cant wait until this is more accessible offline.
Can't wait for this to be open source and on a1111 or something
Can this use my own image or my own video reaembling and looking like me for example doing a music video based on my own selfie images and my own song.
remember when you were at 2k subs?
AI generated videos look like dream sequences haha
Nice lion moonwalk! 😅 But nice to have indeed, very promising!
Pika Lab link for the beta signup is close apparently, sad
Send me a Dm on discord, I'll send you the invite
@@Aitrepreneurcan I get one too? Please?
@@Aitrepreneur Can I get too?
Where Discord?
Thanks Overlord.
Conseguiu?
Question, is there a way to change the aspect ratio, or is there only one option?
Arrg. Useless until I can host it on my machine.
Thanks for update anyways.
I hope soon you can give us the great news of a local running version. Cheers!
nice
Rule 34: Hmmm i wonder what i could do with this... oh, nevermind it is not OpenSource... yet
Awesome!!!
Rats, no NSFW with runwayml (of course it's the first thing I checked, and it was a fairly tame image.).
No way I not use discord because every time I use discord my parents will get ground me! 😠
Hope this will get added to Automatic1111 ^w^
Hope i can run something like this on my pc locally soon.
Hey K, could the ControlNet Reference help with AnimateDiff in Stable Diffusion?
Trippy
and now we just wait for the open source variants that will outperform 😂
I could watch this againanagainanagainanagainanagain....
my take Runaway is 💩(both img2video and text2video) based on the results I got.. SD's animation plugin is still the best option to go but only decent with a video model
I need text to vid and vid to vid for use with AMD GPU - any ideas? Seems like I just have to go back to NVIDIA - not looking forward to it...
Magespace added animated GIFs using images as a reference
9:35 "Add an image, write a prompt or use both"
...
"use both"
so i think you can upload and write the prompt at the same time.
its impossible that a paid services doesn't provide something a free tool does.
Wow! when are we gonna get GEN-2 like open source software to run localy!?
Hi can we use this videos anywhere? no copyright?
It does not give me the option of generate unless i have entered a prompt
Nothing you can't do (better and faster) locally or in a cloud GPU VM.
Instead losing time watching abd hoping that a online service produce something decent, learn how to manage a full AI creative workflow by your own. Is not that difficult.
Understanding how inject a temporalNET is not complicated.
I'd rather wait for a local or one for A111, Gen-2 is nicer than pika but it's as restrictive as CAI treating you like a kid on what you can and can't do AND you have to pay an extreme amount of money just for 2 mins of animation. Then there's pika which may be free but like shown in the video, the quality is terrible and I also rather not show my prompts publicly on a server which all can see.
yey
Is that your real voice?
Woww
naughtiness
I can't wait until there is a working LOCAL solution for this.
Bruh pika closed pretty much instantly after uploading this video D:
I'm going to assume this is not available locally. at least not for AMD users per usual. Kinda cool but still like other video generation methods out there feels limited and Niche. Maybe that's just me from a casual user and professionals will find a use for it. It really has a difficult time with eyes per usual, even stable diffusion gets eyes wrong far more often then anything else usually requiring inpainting over to and regenerate until fixed. it's cool just nothing for me.
Hi Youri
Llm model, this, audio gen, with some python= automated video content creation
Why didn't the runway ml prompt work?
I tried Gen 2 and it just morphs my image into something else entirely, every single time. Disappointing. My guess is too many people complained about a lack of movement and they cranked it too much
They probably cranked the censorship up...
AnimateDiff is not a hoax, but it's insanely hard to install, just take a look at the issues section of the repo.
It requires installing a really specifc combination of torch/xformers/triton
In my experience installing one of the dependencies caused one of the other libraries to break. It's a whole mess that I can't figure out, and I'm not the only one.
I didn't say AnimateDiff is the hoax, I said the fork of it where you could use your own images is
@@Aitrepreneur oh I misheard it, pardonne moi
this application is unsatisfying. I've tried a hundred times and have never had the results shown in this video. I've also tried using photos, but the results are frightening every time.
Either I'm the only one using it badly, or the capacities are inflated.
How to run this locally?
Got a free trail at runwayml. I can't even begin to describe how disappointed I am. The image enhancement tools change my images into something completely different, for example, it changed Nikola Tesla (an old slim man) into a fat & old South American type woman. The movement in the videos is minimal or non existent, it's usually just a tiny bit of zooming in. I can do better in my video editor. I think we may have to wait a few more years before this becomes usable.
when this is possible to do locally then we moving forward.
bruh zero scope xl is awesome
is that how theyre gonna do solo leveling?
I tried several times to get into pika….. no response for weeks 😵💫
The Gen 2 says it can use an image, prompt or both. So they think you can write a prompt for it at least. I'm guessing you discovered it doesn't work?
As I said if you use a prompt in combination with an image, the final video will not look like the base image
@@Aitrepreneur Gotcha.
yeah man i dont know. got a photo of a h0t girl from instagram and told it to make her talk in a video and nothing at all.
The typeform for pika labs is closed
First
We all know what this shit is gonna be used for 😅
*the future is fucking now*
rtx 4060 ti 16Gb Make Sense now
sadly no local version. My A100 is starving^^
Just wait till people start using this for NSFW content
Unfortunately Gen-2 don't work with NSFW pics 😔
Pika Beta is now closed
You can send me a dm on discord, I can send you an invite link ;)
The typeform to join the beta for Pika has been closed?
Send me a dm on discord, I'll send you an invite
Awesome work, but if it isn't open source then its a pass for me. Thanks for the video!
Everyone is talking about local, but the elephant in the room wants to know, when NSFW?
Not until the heat death of the universe.... the narrative against NSFW is getting stronger not weaker....