Це відео не доступне.
Перепрошуємо.
FINALLY HERE! Hedra: Revolutionary AI for Animated Still Images
Вставка
- Опубліковано 15 сер 2024
- hedra.com
Wow! Remember EMO? A video technology we previewed back in February that brought still portraits to life like nothing we’d ever seen before! And then….nothing. And nothing else like it.
Until now.
In this episode, we explore the groundbreaking AI technology known as Hedra, which animates still images by syncing them with audio files.
Demonstrating Hedra's capabilities, the host shows various examples, including different facial expressions and voices, generated from user inputs.
The video also provides a tutorial on how to use Hedra effectively, highlighting its potential in creating lifelike animations.
Note the quick processing times and the subtle realism in the animated faces, making it an exciting tool for creative AI enthusiasts!
00:00 Introducing Hedra: AI Animation Tool
00:54 How to Use Hedra
01:28 Demo: Creating Animated Images
02:35 Examples and Results
03:51 Subscription and Membership Information
04:20 More Examples and Observations
06:25 AI Imperfections and Melting Hands
06:40 Wrestling Champ Story
07:07 Character Voice Experiment
07:36 Steampunk Robot Scene
08:14 Fast Food Zombie
09:48 Poetry and AI
10:53 Emotional AI Narratives
Explore AI Lab Essentials for a deeper dive into these topics with Bob: www.bobdoylemed...
👍 LIKE If you found this video valuable. 🙂
🥰 SHARE If you know someone who might enjoy this video.
⏬ DOWNLOAD or ADD This video to your PLAYLIST for easy access later.
💬 COMMENT Your thoughts and questions are welcome!
📝 SUBSCRIBE / @bobdoylemedia
That's what keeps me going!
🔔 And make sure to hit the NOTIFICATION BELL to stay updated! 🔔
🌎EXPLORE
bobdoylemedia.com
🗓️ MY HISTORY
🧠 meetbobdoyle.com
🤫 Featured Law Of Attraction and Neuroplasticity expert in the book and film, “The Secret”. • The Secret
🎙️ Voice Over Artist 30+ years
📻 Broadcaster / Actor / Creative
❓ASK ME
Got a Question? bobdoyle@me.com
📌 FOLLOW ME
Facebook: / bobdoylemedia
Instagram: / bobdoylemedia
🪙SUPPORT
If you want to support me, the best thing to do is to share the content… sharing is caring!
If you want to support me financially (completely optional and voluntary, but a lot of people have asked for this):
www.thebobdoyl...
Learn more from Bob Doyle Media: www.bobdoylemed...
To clarify something: There is NO time limit to the audio uploads, according to their front page. In the video I wasn't sure. ALSO: apparently this app cannot be used in the states of Illinois, Texas, and Washington.
Yeah, I added that to a pinned comment and should throw it into the description!
We clip to 1 minute right now for the free plan
@@MichaelLingelbach007 Ah, ok. Good to know.
Ownership and Licensing of Your Content
Your Original Content:
Ownership: You retain ownership of any original content you upload to Hedra’s platform, such as an image or video you own.
License Granted to Hedra: When you upload your content, you grant Hedra a broad license to use, modify, and distribute your content. This license includes the ability for Hedra to use your content for training, promotional purposes, and other operational needs of the Service.
Generated Output:
Ownership: The output (e.g., a final video created using your uploaded image) is also considered your content. You retain ownership of the output, but it is subject to the license you grant to Hedra.
@@RiseDefiant thanks for doing that! I don’t feel like that’s so bad.
I just tried it with Eleven-labs script. I now need to go for a long walk over the hills to recover from my output of the download I just played back... I HAVE NO MORE WORDS to describe what this can do.
It is scary!
I started a Sci Fi book about AI 35 years ago. I've gone through 30 iterations and now reality has converged with what I wrote. I'm now making it into a movie. You can't believe how good it could end up.; This would put it over the top.
You do not lose all your rights to the product created using Hedra’s webapp. You retain ownership of both your uploaded content and the final output, but Hedra also has significant rights to use the content. You can monetize the final video on UA-cam, but you must be aware of the licenses and potential complications that might arise from Hedra’s broad rights to use your content.
@@RiseDefiant I think that developments on the subject of intellectual property will not be so good in the future. I always keep in the back of my mind that the content I create using artificial intelligence (and then post on the internet) is practically something created by a machine that BELONGS to a company. Therefore, I think that in the future there will be a big debate (legally) as to whether in the end the creator (and owner of the rights) is the user OR THE MACHINE and by extension the company to which the machine belongs.
keep talkin the talk
@@RiseDefiant I hadn't thought enough about these issues. Thanks for pointing them out.
Nice if you need some soundtracks, I'm here to help
You do not lose all your rights to the product created using Hedra’s webapp. You retain ownership of both your uploaded content and the final output, but Hedra also has significant rights to use the content. You can monetize the final video on UA-cam, but you must be aware of the licenses and potential complications that might arise from Hedra’s broad rights to use your content.
Ok
Michael from Hedra. Great video! Character-2 will resolve arbitrary rendering distances :)
This is great stuff! I’d love to connect with you. Is that possible?
Really appreciate all your hard work, introducing us all those cool new tools I never would have heard about without you mate, Thanks !
Thanks for watching! Really appreciate it!
Long time fan… first time caller. Always look forward to your insight and thoroughly enjoy how quickly you cut through the B.S and get straight to the point. Thank you, Bob.
Thanks for the kind words. I used to be a real rambler before getting to the point. Holdover from my live-streaming days. So I do try to cut any fluff. Thanks for noticing!
@@BobDoyleMedia…and you ditched the distracting music👍
@@BobDoyleMedia…of course, you meant Jennifer Lawrence, no?
@@HMaxTube11 I did! 🤪
@@BobDoyleMedia probably why you got a guy’s image…but “Jessica” shouldn’t have turned out with a man’s image anyway.
I took a selfie pic and a max headroom audio clip and it was perfect even the head movements. I have really bad teeth due to long term health issues so recording myself is less optimal. Hedra is great!
The older guy in the cowboy hat is really impressive. Overall incredibly impressive. Can you imagine what the “good” stuff we can’t get access to can perform?
Oh I imagine it a lot.
No need to imagine. Just look around you.
Imagine this "thing" in a year or two from today. I feel really lucky to have the opportunity of watching history in the making. Awesome things, really.
Try a month :)
It’s nuts really! I don’t think the masses understand yet
@@fonsecarichard I remember years ago a video was published, where the MOVEMENT OF PHOTONS was captured with a camera shooting at 1,000,000 frames per second. In other words, for the first time we could see light moving in a dark space (reflecting etc etc). Few paid attention, although it was a unique ACHIEVEMENT and a great scientific moment. I think the same is happening now. Few realize what it is about, how important it is and what it means for the future. It makes sense of course, time flies, everyday life has problems, it is not easy to pause and watch what is happening OUTSIDE your own microcosm (personally, I happen to be on a long leave from work for family reasons and I have the comfort of being able to to follow developments on the subject)
History in the making is the best way to put it.
@@CinematicSoundMaestroI have been thinking on this A.I. explosion a lot. Like it’s all I think about. I regret to declare that there is only one way this ends, and it ain’t good for us.
Hedra is the biggest breakthrough in A.I since the arrival of Chatgpt4.
Wow, thank you so much for this! I didn't know anything about this app, it just appeared on my YT feed. I've immediately joined their site after watching your video! I'm currently using an AI photo of a 50s singer (I really wanted to use a Creative Commons photo of Eddie Cochran but not allowed unfortunately) - & he's actually 'singing' the lyrics from a still image!! (I had to split it into 30 second clips as it's the free version I'm using). I'm truly amazed at the results - it will feature in the next Eddie Cochran music video!! Many thanks again, I can't stop smiling now 😀
This will be used in the future for moving comics, Action Comics turned into real actions. Also, great for memorials and remembrances. Brilliant!
Now this is what I've been waiting for. I can finally use my generated singer on my generated song, and in between verses I can have the other video creator have the singer move around a bit. Will need a bit of cutting and editing, but will be so much simpler now. Is it weird that I got excited that the background didn't warp while the characters were moving their heads? :)
Update: Just been playing with Hedra... wow, you can get them to sing without even having to separate the vocals from the music, such a time saver... amazing stuff.
As soon as the video ended i went straight to hedra to try this out. WOW! its really is amazing. I am not tech savvy (its just a bit of fun for me) so i have no idea how to use this to my advantage, but to those that do....go get em! Great video😊
Wow! Supercool! Big FANX for sharing, showing and explaining!
My pleasure!
FINALLY! We got one where the body is moving with the head. I don't need to use HeyGen anymore now. I hope somebody creates a local version of this as generating them online is annoying plus I don't want the restrictions. Hopefully, a full body update comes in the future as well.
Wow! Does this mean that the stars in my channel will come to life and start singing? I can't wait to test it :D Thanks Bob!
Holy...This is really insane, thanks for sharing!
Thank you for your wonderful and entertaining tutorial introducing us all to this groundbreaking AI technology!👏🙏
Thanks so much for watching!
11:30 omg, this is astonishingly good 😲😅
The Adam’s apple on that guy with the hat moves! Amazing.
And they swallow and lick their lips sometimes too! (That sound always worse that I mean it.)
Lol exactly! It is really impressive😊
That exponential growth of technology is hitting hard rn
Incredible, I use a wide range of tools including Suno, Pika, Luma, Runway and Elevenlabs. This will bring another tool to my AI arsenal, thanks for sharing.
The guy near the end before u showed paper hat man again was the most expressive and real
I made a blues song in Udio and generated a character and video for it, but sadly I cannot share the video but only the character. My goal was to see if it can sing the blues and how emotional the character would be. I would say it was better than I expected but still some limitations with stronger emotions as you mentioned in the video.
Awesome content Bob! Great to be part of your community!
Awesome!!! Thank you so much for always pointing me in the right direction. Your content is worth its minutes in gold to me. Keep up the great work! Love and light from sunny Khao Lak in Thailand ❤️🇹🇭
Every time I click on a video of yours to watch it, I click like before I even watch the video. In fact, I hit like as soon as it starts. You understand.
That’s extremely generous. Thank you!
@@BobDoyleMedia your content is ALWAYS good. There is always something very interesting to see. Thank you for doing a very good job Bob. Respect.
Ohhhhhhh snap. I am about to use this XD, thank you for sharing.
ABSOLUTELY MAZING AND I HAD NO IDEA, THANK YOU VERY MUCH :)
Wow I'm blown away
This is very impressive stuff. I took a recording of a friend singing an emotional song and a picture of them (he was facing partially way) and it rendered it really well. That being said, I tried it with an instrumental and it did not know what to do with it
Totally amazing. It would be cool if you could insert tags in the text to invoke expressions or pauses. I wonder if it filters celeb faces.
So good to try a tool that actually works as good as the vid said it was - thanks Bob. I can see that in a few months time this will have the potential to bring every character to life in a fully believable way.
Audio may be shortened if longer than 30 sec. is the message I got. Thanks for all your great videos!
BRO! Thank you....thank you thank you thank you. i headed right over while your video ran...and tried it out. AWESOME!!!
Interesting, I was trying another similar tool I found in Pinokio, you can use it locally, but this one looks better. I'll give it a shot. Thanks Bob!
It does better with audio you upload because it has more emotion and natural pauses for breath which makes it seem more natural than the ai voices
This technique was already available three years ago on mobile phones only and very popular at the time
If you’re thinking of Avatarify or something similar , this goes way beyond that. I was all over that technology and even have videos about it on this channel from way back then. This interpretation of emotion is based on the audio itself and is much more sophisticated. But to your point, I had a ton of fun with that tech back then and reviewed a lot of them. If you’re talking about something else, what was it? Super curious!
ua-cam.com/users/livety9hYLEHdMw?si=K8ka0mq3swlHEweH
Not even close. The ones that were out years ago just warped the image to make it look like they were talking. This is totally different.
I am in this AI industry and yet, this totally caught me off guard. I am really impressed of the range of emotions it can reflect and (for God's sake!) Adam's apple of that man is moving as the way it should 😂 Thank you for bringing this marvelous gem to our attention 🙏
I just tried a little clip to help in bringing a music video to life. Wow!
Hai…❗️Bob Doyle 💪
This is quite an amazing ai … I pray you reach out for more of this 🙏
The Hedra AI app has been blocked in Texas due to privacy concerns and regulatory compliance issues. The Texas state government cited that the app did not meet the necessary data privacy standards required under Texas law, particularly regarding the handling of user data and the transparency of data usage policies. Additionally, the app faced scrutiny over its methods of generating and storing sensitive information, which state regulators deemed insufficiently secure and potentially vulnerable to misuse.
Texas govt is so backwater stupid. It's run by a pack of corrupt individuals who are supported by an equally corrupt legislature and judicial system.
This is excellent , Ive been using all types of AI facemove apps and programs and what i find they all have in common is that the face of the person's emotion has to match the text to make it work good. I love how this program AI captures the emotions of the audio so all that is needed is the right pic to use , like if you use a happy audio , you need to use a pic of the person smiling etc. One thing it can use as a improvement is that the motions need to be smoother like other apps. I think using this program I can fix that by taking the video and put it to 60 frames per second , that might work or the other solution is for the AI program needs to fix it in the software. Keep up the good work , you make great videos
Our model supports 15, 30, and 60 fps but we only output 30 in the current UI. We'll add more advanced options in a subsequent release
@@MichaelLingelbach007 if you can output to 60 frames that would be good but even better if you can make the model more smoother in motion it would be a vast improvement. I used it lot right no and I notice the AI likes to close its mouth a certain way like a frown on pauses. It’s good but it tends to be repitive and it has a form of a frown that you don’t want say in many type of emotions , especially happy. What I’m saying here is if you can have a option to tone down the emotion and make different variables of the emotions then that would be a vast improvement , regardless keep up the great work on this app
@@danielbanic3738 many of the improvements around video quality are just a matter of time and scale, in progress. We’re adding more controls. Model can output 15, 30, 60 fps. 30 is the default.
@@MichaelLingelbach007 Just a little advice coming from a artist in music and video from a artistic point of view , your AI does'nt need to have a upscale of the video , it looks natural just the way it is , in fact if you upscale the image video it will look artificial. The upscale is only good for people who wat to do AI art but for natural pictures it looks best with no upscale , Keep up the good work
@@danielbanic3738 duly noted and 🫡
Testing now. Fantastic concept... but way too low resolution to be used in real word. Strangely in the home page they said:"free and paid plans"... but I dont see anything about, till now...
This is quite exciting -we are getting near to making real narratives with AI actors.I can see this sort of tech being used in a good way and a not so good way.
Wow! This is truly AMAZING!
Bob, have you not heard of the likes of D-ID and HeyGen that can animate photos? These have been around for a couple of years now. Thanks for the reminder about Hedra.
Oh, yes. Very familiar. Have been on the hunt for open source alternatives for all that time. And while this isn't open source, it's currently free, and I think WAY more expressive than either of those two solutions, at least right now.
I noticed that it decreases the quality of finished video. You will need upscaler for it.
Running some upscaling right now.
These are wonderful, but my only complaint, is that the moving ones, don't look as clear as the still ones. There is an element of blurriness, and one female with freckles, showed her blurry and without freckles. But that last one you did with the man in the paper hat, I thought I saw that cook in the background move just a tad, take a look and see. very very impressive. Can you Use A real picture of a real person to do the same thing, but with a disclaimer saying it's not the real star?? Thanks Bob. Terry
Yes this is a great site, so far for me it's the best., I hope it stays free. Thanks for the video. Why the hell can it not be used in those states? Did they cheat on their exams? SUBBED.
Great 👍 got my ideas flowing Bob
Awesome video. I just made a music video earlier and the lip syncing wasn’t that good, so I used lights and special effects to distract you from the lip syncing lol. Works pretty well actually but I like this solution even better
Wow just so innovative and outrageous as well we must say.
Agreed!
@@BobDoyleMedia I'm currently looking at the video and have subscribed to your channel as well.
I have a certain feeling that AI will never be perfect, and I hope not, I hope that we will always be able to see that it is AI, one reason being that we are human and we created AI not the other way round.
yes, I upload another one, looks greate , hope we can do full width as well
Most probably in the pipeline to do, but yep deffo 1X1 at the moment.
Love it and perfect for what I am working on.
Love IT! I was just looking for a lip sync AI and here one is
This feels like just a demonstration without any practical application you can’t make your characters say anything bad and develop a story especially if you want a bad guy threatening the good guy
And yes even a beta should have a a purpose or at least a end goal
@@anthonystutsman2591 Our moderation settings are currently too sensitive, we're dialing it back. We ran a 1 week alpha before release and it was not flagged to us, we apologize for the inconvenience.
Been looking for this!!!😉
Love the channel and videos! Thank you Bob Doyle!
I will give this a try. Looks like it could be fun. Thanks 4 sharing. 😊
I’m guessing it is pronounced Head-draw since it is drawing Heads. I can’t wait for their future project called Butt-dra! ;)
How does it do when feed a song? Can it sing the lyrics?
Wow this is mind blowing.
You do not lose all your rights to the product created using Hedra’s webapp. You retain ownership of both your uploaded content and the final output, but Hedra also has significant rights to use the content. You can monetize the final video on UA-cam, but you must be aware of the licenses and potential complications that might arise from Hedra’s broad rights to use your content.
thanks bob, all the best things in life are free
Thanks
Thank you Bob.
Hi Bob
When u make videos for you tube it asks if u have used AI I believe, when u go to upload them...does that mean you cannot monetize if u use AI ?
Great demo Bob thank you
Wonderful Video Mate 👍
I can see how clever the tech is and why people get excited about its realism but I haven't heard many people say what the point of it is. What are good uses for this? Or is it just a bit of novelty fun for a few minutes?
Good for people who prefer to use avatars instead of their faces for UA-cam content for any reason.
it's fabulous and amazing
At 6:45 I was like "What!!!!"
thanks for this - great insight and research - ya just saved me a ton of time - thanks
Great video. Thanks for sharing.
I remember when software like Crazytalk (by Real Illusion) came first in 2010's en then I see this. I got vision that Hedra start it all and will becom like next tech will come as Crazytalk became Crazytalk Animator next: Full boddy and head AI realistic animation and multi caracters and scenes complete AI software.
Not as good as VASA-1 but way better than anything else I've seen and its a contender, I'm sure it will improve.
That's really great. But what we really need, is finding a way exporting this to a system that can transfer text to voice in milliseconds.
AI lord is here!
You got excellent separation of landscape and sky. Does not look like many floaters. Did you have to run high iterations to get that? How many images did uou use? Thanks for uploading
You’ve got a great voice and do good voice characters, Bob. Have you done professional vo in your career? You sound similar to Mike Rowe of “Dirty Jobs” fame.
Yes. 40+ years in VO. Bobdoylevoice.com 👍
I really need to figure out how I feel about all this. I have an audio from 1976 of all my family members just carrying on regular conversations. I put it on a CD about 15 years ago. Technically, all my deceased relatives could "come to life" again with this. But do I even want or need that? No matter how good a fugazi is.... it's still a fugazi.
@@JB-js4xi What if we’re ALL a fugazi already?
@@BobDoyleMedia well I suppose to a degree we are since we represent our selves differently to and for different people. So which part of us IS real?
So, I gave it a go. I applied Hedra to a drawing that I had made and utilised a character that Hedra created to my simple description. Here's the short AI film interview that is the result: ua-cam.com/video/i5_uxx6ERRY/v-deo.html
I want to test Hedra, but it has a strange behavior: It requires you to link your Google account to it. No other options. Why? I prefer to create an account from scratch.
Can't wait to try this
Super interesting. Found this product far better than HeyGen or even Yepic which we've used both. For Season 3 we're thinking of bringing our hosts BACK as Avatars. Stay tuned and please take a look at our AI news and information over on our channel. Thanks!
QUESTION: Is there a way to make only SUBTLE head movements? Like lets say we have a character who is gentle and humble. Can we choose amount of movement? Do you know of any program that can do that? Can you keep an eye out for this for me? l0l :) You're awesome! :)
Brilliant ❤thank you. Loved thre end bit😂😂😂😂
What were the dimensions you of the photos you uploaded or did you only try AI images? Not sure if you mentioned this.
What type of file output options do they offer?
This is mindblowing
Well. Done 👍
Great stuff Bob .. I now see dead authors, poets, writers, lecturers etc coming to life and reading us their works ... in their own voices ... if there are any clips of that around somewhere online. Anyone looking to start a new channel ... get busy.
It's definitely being done. I'm seeing a lot of AI content (most certainly unauthorized) using clones of various thought leaders, authors, and speakers. I'm only ok with it if they clearly say it's AI. I did one with Neville Goddard as a test quite some time ago, but was clear about it being AI. ua-cam.com/video/AuUc3m4eBuY/v-deo.htmlsi=KK7pef_gmVCjKieM
Man that's crazy, but you should read the agreements you accepted before using it, read it very carefully the first point: By checking this box, I hereby represent that (i) no image, video or other media I upload will contain images of any individuals other than myself, and (ii) I have read, understand and consent to the Hedra Privacy Policy, Hedra Terms of Use, and Hedra Acceptable Use Policy.
Yeah, I think they make that fairly clear up front. I seem to remember "agreeing" to that upon signing up. I generally use the in-app face generator anyway because I'm mostly using this for character work rather than real people.
@@BobDoyleMedia I think it can be more to prevent deepfakes than preventing people from creating from generated faces, because in their featured part, there are examples of not real people
@@BobDoyleMedia I wanted to try with ai generated characters but I was afraid about that
This is the best by farrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr!
What do you use for the background to record? So it picks up you, you're web browsing background
Good one, many thanks.
Good video. Would have liked to see A LOT more examples of it singing, though. 🤷
(as of 7-2024) Hedra is not available to us in TX. Bummer!
@@PrayingMamma2 They are working on that…👍