These AI Filmmaking Tools are Incredible!

there's a new app that creates incredible VFX shots using just your phone we came across some Sora Updates this last week that absolutely blew our minds and we're leaving to the can International Film Festival to share the future of film making with the industry welcome to AI film news I want to kick things off by showing you a really interesting tool that we came across called simulan that basically allows you to shoot an augmented reality version of a scene and then upres it using artificial intelligence to get maximum quality basically the way it works you use lar technology to scan a 360 scene to get all of the reflection maps and the data that you would need for the lighting sources then you go in and you use your phone to shoot a very simple representation of your character in that scene so it uses the AR features on your phone to composite the character in the scene and you can move move your camera around but it looks kind of low res directly from your phone but that's where the magic happens because using this technology you then send it to the cloud from your phone and it comes back a few minutes later with a highly realistic composite with that 3D asset so it's basically this AR virtual production workflow that creates some pretty interesting results and there have been some really cool demos online here's a few examples so we have this shot of this robot car character that basically is doing character replacement jumping over a bridge and it looks pretty dang realistic that compositing work looks borderline cinematic we also have this scene of two robots fighting in a garage and the camera shake the tracking looks pretty realistic now the shadowing isn't 100% accurate but it's not too bad especially considering this was shot and entirely composited on an iPhone we also came across this scene of this Mech suit robot walking through a park and the reason why this is so interesting look at the shadowing so when the character walks through the sun it has that direct sunlight but when it walks under the shadow of the tree the light pours through and it actually looks realistic and if a VFX artist was to put that together it would be a lot of work to simulate that lighting to match and so the fact that this app was able to do that so accurately is really mind-blowing and finally we have the shot of a Porsche here and just look at the reflections as you get closer to the car the reflections move around in a very realistic way and as you get really close up you can see the details you can see the specs from the paint on the car and again this is entirely taking place in the cloud so this is a phone to a cloud compositing workflow I think when tools like this are paired with some motion tracking tools and actor tracking abilities you can pretty much imagine that pretty soon you'll be able to have some different characters in your scene whether they're monsters or robots or any other character that you can imagine very easily and the compositing process will be way easier than the current VFX pipeline the team at Mid Journey also released a public beta to anyone that has generated over a 100 images if you use mid Journey consistently there's a good chance you now have the ability to use the tool online all you have to do is go to Mid journey.com they also have a really interesting feature that they just released called Rooms basically this is very similar to a Discord thread where you can click in here and you can basically see other people generating prompts in a specific style and you have the ability to enter into a voice chat if you want to chat with other people I haven't really done that yet but it's there if you're interested we also saw a really interesting Sora demo this last week somebody prompted a man in a gray suit basically walking down an Alleyway but then they decided they wanted to change it to a woman so they just changed the prompt to a woman and we now have a new character they also changed it to a six-year-old man with a white beard it changed it to that man and then a metal industrial robot so this is pretty cool because it basically shows that you have the ability to pull seeds inside of Sora now of course we don't have access to Sora right now to fully test this out for ourselves but it is cool to see that this functionality is coming very soon and while we're talking about Sora we came across a video this last week that blew our minds basically Paul trillo who has been creating the coolest Sora content put together a music video for the band washed out and I actually was hanging out with Paul yesterday at the Cog X event out here in Los Angeles and he was talking about the process of creating this music video it's very experimental but it's really incredible to see how each one of the scenes really blend into the next one I think he did a really good job with curating the shots and pulling something off like this using previous technology would have been incredibly cost prohibitive and very challenging for most Productions and so the fact that you can type in a prompt and get the visuals and the storytelling possibilities using just prompts alone is really incredible and speaking of incredible we came across a really cool tool that you can play around with right now that basically allows you to change the clothes of the actors inside of your images so you can change their outfits based on uploaded images this of course is going to have huge implications for the future of advertising and film making so let me show you how to use this tool basically what I'm going to do I'm going to leave this sweater as the reference refence image because I want to composite this sweater onto a different person and I'm going to go down here to our human IMG file section and we'll clear the file and then we'll click the upload button and we will double click to upload I have this image of this woman and we want her to wear the sweater so we'll go ahead and open that and you can of course go down here and adjust some of the settings there's a lot of them it can be quite confusing uh but for this specific image we want to change the upper body because we're only changing the sweater and not the entire outfit and we'll go ahead and click run and after just a few seconds it generated this image here and got to say it looks pretty darn good obviously it's not photo realistic to the original uploaded reference image but it's pretty good and I think at a quick glance you never would know the difference so of course I'm curious if it does a good job with a photo of me so let's go down here and add it in I have this photo of me so we'll go ahead and open that okay let's take a look at the image and actually it did a pretty good job I think if you really look at the fingers there the fingers are like way too long but the overall photo it's pretty darn accurate and I really like how it's not reinterpreting the face it's still the same face but it changed the clothing and definitely at a quick glance I would think that this is a realistic photo and I'd want to show you how you can also change the clothing so all you have to do is go to the Garment section here we'll clear that file and click on the upload button and I have this image of a super expensive Louis Vuitton shirt that I would never buy and let's go ahead and upload that and run it okay so as you can see it did a pretty good job obviously it's not the exact same t-shirt so this technology still has room to improve but it did a pretty good job especially considering that the shirt that we uploaded was pretty complex and it's pretty weird compared to a normal shirt I think that this type of Technology really illustrates that in the very near future we will be able to change the clothing and the costumes of virtually any photo or video with just drag and drop and while we're talking about fashion we saw a really interesting story from the Met Gala basically open AI teamed up with the Met to create an AI chatbot of Natalie Potter who was a New York socialite and basically wore a fancy dress a long time ago it's also really interesting to see that open ai's CTO Mera Madi basically attended the Met Gala and I think it really showcases that traditional art and fashion institutions are beginning to embrace AI in their workflows we also came across a really interesting lighting transfer tool that basically allows you to change the overall lighting of any image and it does a pretty dang good good job so I have this photo of this woman you can see that the lighting is really flat and I want it to be more Dynamic so what we're going to do is bring this image into the tool and now you have to type in a prompt for our prompt I'm going to say woman cinematic lighting with a gray background just basically describing our scene and you can see that we actually have some options here to change the direction of the lighting so we can have the light come from above we can have it come from below let's go ahead and do top lighting and go ahead and click relight okay so within just a few seconds it created this image here you can see that it did a really good job actually there's a lot of texturing on the skin the lighting looks realistic you can see that there's Shadows coming through that look really good and I think it really does showcase that in the future we will be able to relight virtually any image or video using a tool like this there's already some AI relighting Tools in Video Edit applications like Da Vinci resolve but a more Dynamic relighting interface is likely just around the corner and speaking of around the corner there are a ton of AI film events that are happening the first one that I want to let you know about that we are super excited for is the can International Film Festival the team at curious Refuge is teaming up with Leonardo to sponsor an AI film making Meetup at the can Innovation Studio it will happen on Friday May 17th from 5: to 10: p.m. you can come out to the event for free drinks we're going to have some really interesting AI Tech demos and some AI games I should note that you need a badge for the event but if you're going to be there we would love to see you we're also giving four talks in collaboration with can next and Microsoft across multiple days you'll find links and more information below this video Shelby and I will be immediately leaving can to give a two-day workshop at the national Film School in Denmark the event unfortunately is sold out but they have opened up a wait list for folks who may be interested in attending I should also note that AI on the lot is happening in Los Angeles on May 16th this is one of the coolest AI film making events in the entire world over 500 people will be in attendance and they're going to have panels and discussions with some of the most talented Ai filmmakers and creative innovators in the world they'll have speakers from open AI the director of Black Mirror bander snatch will be there and the virtual production supervisor of Fallout and rebel Moon you can use the code curious at checkout to get 20% off your ticket we're super sad that we won't be able to be there but if you're in the Los Angeles area I highly recommend attending the event and speaking of events we'd like to extend end a thank you to the Cog X Festival in Los Angeles Shelby had the privilege of speaking on a panel about AI tools and we had an incredible time learning about the future of entertainment law and so many important Concepts at that event there was this really cool demo where they had a piano player play three different pieces of music one of them was an original Symphony by chopan the second was an AI generated Symphony and the third was created by the artist and then they had the audience guess which one was AI generated and it actually was really challenging for the audience about a third of the people actually guessed it correctly submissions are also open for our AI film making competition again we're giving away an apple Vision Pro to the winning film trailer all you have to do is go over to our website you can download the rules and then upload your trailer we already have 73 submissions but we imagine that that will be even more in the coming weeks so you still have plenty of time to turn in an awesome submission and I should note that you can turn in more than one trailer so if you have multiple trailers that you think could make the cut I highly recommend uploading them to the platform the government of Ukraine is actually using an AI spokesperson to get official messaging out there it's really interesting basically they now have an AI foreign Ministry spokesperson called Victoria shei and basically she will say words and communicate information that is written and verified by humans in the Ukrainian government but basically this is helping them save a lot of resources that are obviously challenging to have during the crisis that's unfolding there in order to fight misinformation they're also including a QR code with all of their messaging that will directly link to the government's website the team at udio also released a few really interesting updates to their tool that are really interesting the first update is giving you the ability to extend a track for up to 2 minutes so up to this point you could only extend a track for 30 seconds now it can be extended to 2 minutes the second is giving you the ability to do audio in painting so for example in this video we have this example of a track let's go ahead and listen to [Music] it warm true [Music] never but what if you want to change the lyrics in one specific spot well now you have the ability to go in and inpaint that specific area it's really cool so they have in the demo where they change the words and here is the [Music] result feeling never when I with all right so you can see it actually did a pretty good job there's also a really interesting demo where they showcase adding in a piano solo to this track and I think it really does showcase that the future of music production is going to have audio in painting as a very essential tool and speaking of essential tools we had some more information from Sam Alman about chat GPT 5 this last week basically he said that chat GPT 5 is going to be a lot smarter than chat G bt4 which makes sense but what's also interesting is that the team at open AI is actually working on a Google search competitor that could be out very soon in fact by the time this video launches this new search feature may be available to the masses this of course is a very real competitor to Google and in the near future we could be realistically picking between using a chat gbt search experience and a Google search exper experience and the team at open AI also came out with a tool called chat GPT classifier that basically has the ability to go in and identify content that was created using Dolly so it basically is in charge of identifying content that was created using AI versus content that is authentically real and it's actually pretty accurate it says that it has about a 95% accuracy which is pretty good but the problem is if you add just a little bit of grain to your images the accuracy goes down by a lot so of course using AI to identify fake content or AI generated images or video is really great but of course it will always be this back and forth process to upgrade the tools that can identify AI generated content and the tools that are trying to mask it Spotify has also tested out using AI to translate their podcast and it's actually pretty interesting so you can click the link below this video to go to the voice translations page and the thing that makes this really interesting is the fact that the voices actually are in the native tongue of the person talking so for example they have armchair expert here with Kristen Bell and Kristen Bell and Dak Shepard they sound like themselves but just translated into Spanish more than likely by the time the end of the year is here most content that you consume online will be in your native tongue and of course this is going to open up so much reach for people that may have not had their content found in their original language the team at P laab also came out with the ability to add style presets to your Generations inside of their tool let me show you how to use it all you have to do is go over to the paa website and you can simply generate whatever you want you can go down to the prompt here type in your prompt and generate your video I don't want to generate my own I'm just going to use one of the videos from their explore page here so we have this video of this sizzling bacon looks delicious but let's say that we actually want to change this into a different style all you have to do is click the edit button and you'll see that it pops it up down here in your prompt bar and now all you have to do is go to the style button and you can see that there are the style presets so there's all sorts of different options that you can select let's go ahead and click the claymation version I've also found that it's very important to prompt in what is inside of your scene so we'll say a pan with bacon and go ahead and click generate so let's take a look at our results so first off we have claymation here and I love how it put googly eyes on the edge of the pan like claymation definitely has googly eyes which is fantastic we also have 3D render here which looks kind of photorealistic I don't know if I'd say it looks quite 3D and I love how the bacon has turned into Canadian bacon and it's also levitating off the Pan super funny and the last one we have here is this anime style and it actually look looks pretty dang good I'd say if you told me that this was just from an anime I would totally believe you because it looks great and that brings us to our AI films of the week the first film that I want to shout out is of course Paul trillo's music video we've already talked about it but I think it's really interesting that it also received a staff pick from Vimeo it seems like AI generated film concepts are getting more accolades and really showcasing their ability to tell new stories next up is an advertisement from Paul Montreal which is just a fantastic name Paul basically put together a quick advertisement for a 1980s inspired Diet Coke commercial where some women take some time off from their work and let's just say it's not entirely about the diet coke he put the project together using let's AI for the visuals runway for the video 11 labs for the voice and he used some music from art list and our final project of the week comes from IIA Shopco who put together a film called the rainforest it's basically an Adventure film and it's a really interesting concept because it started with some vintage video that looks really darn good and then it evolves to take on this sci-fi tone I think is a really compelling story would love to see a feature film version of this project and the last story that I want to leave you with comes from the world of country music basically Randy Travis had a stroke and it left him unable able to sing but using AI a team was able to train a model on his voice so that he could sing a song again in an interview with ry's wife and the music producers that actually put this together the team said that they actually hope that the song will work to educate people on the good that AI can do and not the fraudulent activities that so frequently make headlines so I think this is a really good example of how AI is making the world a better place thank you so much for for watching this episode of AI film news of course if you're going to be at can please come by our AI film making meetup on Friday the 17th you can of course subscribe over on Curious Refuge to get AI film news delivered directly to your inbox whenever you do you'll actually get a 5day intro to AI film making course that's super awesome and of course you can like And subscribe here on YouTube if you want to get more AI tutorials and news directly on the platform thank you so much for being an incredible member of the Curious Refuge Community if you ever have any AI news or insights that you'd like to share with us be sure to let us know in the comments we'll see you next time

Channel: Curious Refuge

Views: 87,445

Keywords: SORA, openai, sora news, sora, ai videos, image to video, ai movies, text to video ai free, midjourney, midjourney v5 tutorial, midjourney multiple characters, ai generated art, adobe firefly, sora ai, openaisora, sora release, paul trillo, curious refuge, ai films, ai short film, hyper realistic ai video, ai trailer, ai, artificial intelligence, ai film news, chat gpt, caleb ward ai, generative ai, midjourney video, gpt 4, ai news, AI VFX, vfx, ai vfx tools

Id: V_UQCrPbfJE

Length: 21min 45sec (1305 seconds)

Published: Fri May 10 2024