Nikola Todorovic

SD3, the newest AI image generation model by Stability AI, is out! How does it hold up? One especially cool thing: it's pretty great at text! Check out our Storybook Studios experiment to improve the results right out of the box:

11

4 Comments

2D noise to 2D #omage (or "Object-iMAGE") a way to describe a 3D patch with a 2D image) + texture, to 3D. A new approach to txt-to-3D #GenAI... Interesting 🤔 #2Dto3D

5

1 Comment

AI is not coming for your job. ❌ People who use AI with their traditional skill sets are coming for your job. 😱 At Sombra Labs 🧪  and ALONGSIDE/ Global we believe in merging the worlds of software development with VFX artistry. 🤝 Machine learning combined with the VFX expert arsenal of tools such as Autodesk Flame, Foundry Nuke, Adobe After Effects, and so many more is what will elevate us to the next renaissance of content creation. Here we used SAM2 and Deepmake to recolor our latest Runway Gen 3 test of the amazing Porsche AG Targa 4S in only an hour. We went from green 💚 to ❤️ red seamlessly using AI alpha output combined with Autodesk Flame. Bespoke AI / VFX pipelines are coming and they are coming fast. These tools allow us to color grade, composite and do so many more things 5x faster with 5x the control over segmentation. Exciting days ahead. #ai #genai #runway #vfx

323

44 Comments

🚀 Luma Dream Machine: A Game Changer for AI Animation! 🎬✨ Check out these two animated clips I created using Luma Dream Machine. I did a quick audio and sound effects in post-production, but all the animation magic was brought to life with Luma. Images were created in Midjourney. #Animation #LumaAI #Innovation #PostProduction #AudioDesign #GameChanger #CreativeTech #DigitalArt #AnimationIndustry. #LumaDreamMachine #AIAnimation #GameChanger #soracompetitor #immix #midjourney

23

5 Comments

Here's a look behind the scenes of how I used a 3d model, generated from an image, as a "driver" for ai animation. In the near future we'll see more 3d tools that support these types of workflows, with much more emphasis on enhancing user input instead of just letting the ai do everything. I used the 3d model to generate keyframes that I then animated in Luma Dream. By using a 3d model with a diffusion "layer" done with Krea, I got quite high quality frames with a relatively high consistency as well since the AI didn't have to "hallucinate" that much. I upscaled and detailed the frames using Magnific. Using semi-detailed 3d models as drivers for gen ai is very powerful, and it allows us to use models that are sculpted in a more free-flowing type of workflow that doesn't rely so much on high surface detailing or very time-consuming finish, but instead relies more on gestural 3d sculpting. #art #ai

2,068

124 Comments

Not far from some #ControlNet approach of txt-to-img #genAI tools meeting NVIDIA's #GauGAN, here is a way to attribute semantic sense to colors and paint them. Usable online and locally. #genai #inpainting #streammultidiffusion #txt2img

4

If you can get past the high queue times 😅 what Luma AI Dream Machine does exceptionally well is combine 3D spatial data with AI interpretation - meaning it does an excellent job with consistency and logic. These are just some initial tests but its clear that the team is really on to something here. I'm not sure it quite kills OpenAI's Sora, but its definitely a tremendous step above Runway and Pika so far. As always I wanted to test it with animation as well as live action, and while the results are definitely hit and miss its a very interesting entry into the text/image to video race. What are your impressions with using it, and have you found any interesting tests yet? #lumai #generativeai #ai #videoproduction #animation #openai #sora #dreammachine

27

16 Comments

Luma AI just released new feature. Give 2 image references - transition between them - get this 😮.

28

6 Comments

One of the main challenges in the creative landscape is choosing the right tools that fit your needs as an artist or organization. With new AI tools being released frequently, selecting the right ones can feel like solving a complex puzzle. While AI tools often seem easy to use, getting the most out of them requires extensive knowledge and strong communication skills. Artists need an understanding of color, composition, perspective, and lighting to guide AI tools effectively and evaluate the results. Whether you're an individual or an organization, staying updated with the latest AI tools is crucial for delivering high-quality creative output. We started testing Midjourney early on and, after creating close to 45,000 images, it has become an integral part of our creative toolset. Our Visual Art development team enjoys using it, enhancing our storytelling capabilities. Recently, we tested Luma AI Dream Machine, and it shows great promise for our toolset. Check out the test results below using Midjourney and Luma AI Dream Machine. Midjourney Luma AI #lumaai #dreammachine #midjourney #lumaaidreammachine #ai #aicontent #prakashkumararajan #wildvfx #wildvisualeffects #imagetovideo #aivfx #aianimation #midjourneyai

21

1 Comment

I did a light comparison between Luma Dream Machine (including the enhanced version) and Runway Gen-3. Three different styles of text prompts and 10-second videos based on them. Here are the first clips produced by the applications. From these, it's clear that it's still quite a gamble (although Luma has the ability to animate from images, making the results less random). And since these are developing so rapidly, I guess we should have much better control by the end of the summer. PROMPT 1: "Hand held camera shot, chicken turns to egg, reverse transformation, educational video" Runway is the clear winner here, and even though I don't like the crossfade, the concept matches the prompt. Luma Enhanced understands the handheld camera best, but it creatively misunderstands the reverse part of the prompt, making the rooster (!) walk backward, totally without the transformation. As a fitting bonus for an educational video, Luma Enhanced offers subtitles — probably in some invented language (or does anyone recognize it?). The basic version of Luma is completely off. A roughly seven-fingered clay hand seems to be cooking eggs, with no sign of a chicken. PROMPT 2: "Drone shot, animation, Bill is sitting on a prison roof and throwing potatoes down, Italian 1960s style cartoon" Runway wins the second round by a landslide, although it didn't understand Bill as a person on the roof but wrote it as text instead. However, the style is captured very well, and the potato is flying. Neither version of Luma managed to create a cartoon look. Luma Enhanced ended up with some sort of CGI graphics, while the basic version is purely photorealistic. Luma Enhanced scores points for the nice camera movement, and both Luma versions captured the drone prompt well. It's hard to say what's actually on the roof, though. PROMPT 3: "Moving dolly camera on a track, cinematic action scene, SWAT team handling a hostage situation in a toy factory, criminal wears a pink furry costume" With the cinematic prompt, the results are more balanced. No video shows the hostage, but otherwise, the elements are captured fairly well, and Runway's location most closely resembles a toy factory. The basic version of Luma and Runway could be shots with a dolly camera, while Luma Enhanced seems to be handheld (which is nice for an action scene though). The basic version of Luma makes the most sense, clearly showing the police catching the criminal and leading him away. Luma Enhanced goes completely off the rails, and it's hard to understand what's happening as the characters are morphing. Runway's beginning looks promising, but it would be nice if the SWAT team reacted somehow to the pink bear. Maybe they just don't understand it's the criminal. :) That was today's test. Something different next time. Which camp are you in, by the way? Do you use both or something else entirely?

24

7 Comments

Preparing with Yu-Jun Yeh for the live Intro to AR workshop we’re giving at #UnrealFest and wanted to check the latest on Android vs iOS. Android has better image tracking, but other than that, just about everything else is as good or better in iOS, and the workflow is very straightforward as well. What do you think? Have you tried to get an Unreal project into mobile AR before ?

49

13 Comments

A quick guide to choose the right #SDK for your #XR project.

3

Game Cinematic Experience Leveraging the power of Luma AI and Midjourney, A creative journey to produce captivating generative videos. Utilizing Midjourney for imagery especially the 2x option and the Luma Dream Machine, I crafted a compelling storyboard with a cinematic feel where you reveal the story on 2nd half just like in movie and game teasers, off course there are glitches which i know = ) -- while i was working on a Hunter game, i created this trailer to reflecting the gaming industry's dynamic and immersive nature. we are waiting for SORA AI, but i think Dream machine just filled the craving for a while. This Project took 20 hours -  ------ #DreamMachine #Luma #AiVideo #Game #midjourney

76

16 Comments

Check out Tony Lyons Latest Stamps Workflow with Deep Renders you will learn so much about Script Management, Optimization etc. it's a really great tutorial Highly Recommend to Watch. Thanks for the Amazing Tutorial Tony Lyons #nuke #compositing #deep #workflow

41

Want to boost your efficiency? 🔧 Creating custom tools is one way to go (or buying one if you're short on time). I just launched a Houdini tool that makes it super easy for objects – like these Pringle chips – to follow a path! You can grab Trace Curves right here 👇: https://buff.ly/4dk3FkL #Houdini #3D #Tool #Curves

743

16 Comments

Playing with the new keyframe feature from Luma Dream Machine! It was incredibly fun and intuitive to create this anime clip. Highly recommend exploring it! Animation: Luma Images: Midjourney and Leonardo Post Production: Capcut #AIAnimation #LumaDreamMachine #KeyframeAnimation #AnimeCreation #DigitalArt #TechInnovation #CreativeTools #AnimationCommunity #immix

10

🚀 AutoVFX: One-Click VFX Creation? Here's an intriguing project by a group of students at the University of Illinois. Imagine a tool that could be described as a one-button Houdini or Maya—complete with a sprinkle of Nuke for good measure. The concept is as ambitious as it is bold: 1️⃣ Upload a video. 2️⃣ Speak into the mic and describe what effects you'd like to apply or how you'd like objects to explode in the scene. 3️⃣ The system responds with a "Got it!" and delivers photorealistic, dynamic VFX with no manual effort. Here’s their vision in their own words: “We present AutoVFX, a framework that automatically generates realistic and dynamic VFX videos from a single video and natural language instructions. By leveraging neural 3D scene modeling, LLM-based code generation, and physics simulation, AutoVFX enables physically accurate, photorealistic effects that can be controlled directly through natural language commands.” Ambitious much? 👉 Check out their demo here: https://lnkd.in/gdGXheZX What’s fascinating is the ingenious stack of open-source tools and frameworks they’ve pulled together to make this work: - Gaussian Splatting submodules - diff-gaussian-rasterization - segmentation & tracking modules - Grounded-Segment-Anything and GroundingDINO - recognize-anything - lighting estimation modules - PyTorch3D  - Trimesh  - Tracking-Anything-with-DEVA - Blender for good measure (because why not?) - BakedSDF in SDFStudio for surface reconstruction with custom tweaks. Sure, it feels like a proof-of-concept or a graduate thesis project, but their courage to tackle such a massive idea is awe-inspiring. 📌 For those interested in the nitty-gritty details, check out their GitHub: https://lnkd.in/g7g2QBbt This might not replace your favorite VFX tools just yet, but it’s certainly a glimpse into the future of accessible, automated content creation. What are your thoughts on projects like this? #VFX #GenAI #AI #Automation

47

Sound On!🔊  I’m absolutely stoked and honored to be one of the lucky few to join Luma AI’s new Dream Machine Creators Program! 🤘😎🤘 Midjourney + Luma Dream Machine reel Vol.1 😍 I’ve been a Pro Luma member for months now so let me quickly break down the features that make Luma’s Dream Machine video generator so special to me: ⚡️ Industry’s Fastest Generation Time ↳ You can literally watch the progress bar speeding across the thumbnail, I think Runway is the runner up in gen speed.    🖼️ Text-to-video + Image-to-video  ↳ Text-to-video is of course the OG video gen trigger, with image-to-video now quickly becoming the standard, and v2v coming online as well. Control, control and more control! 🎞️ First Frame + Last Frame Animation  ↳ Luma was the first video gen platform to release this feature (look it up!) and it caught on quick. It’s super powerful and absolutely necessary for better control over the output of any video generator 🧵 5s Video Extend  ↳ Extend your generations 5s at a time starting with the ending frame. I think the max is extend on a single gen is 25s, but you can always start another scene by extracting the final frame and using it to seed your next video! Also, you can input a prompt for the extend (or leave blank), create a 10s seamless loop, and choose with or without prompt enhancement. No new image/keyframe control yet for the extend, but I expect that to come online sometime soon! ✨ Prompt Enhancement on/off ↳ Tick the Prompt Enhancement checkbox to let the internal prompt engineers sprinkle some back-end magic on your prompt. This also means you can tick the box and leave the prompt window blank to see what the Dream Machine does with only your frame image(s) as the input, and it’s pretty awesome! ♾️ Video loop on/off  ↳ My start in motion graphics was in the early days of GIFs and Cinemagraphs, so seamless looping animations just make me happy, and I can actually turn them into GIFs super easily with tools like Adobe Express and other online freebies 🎥 Camera Motion Control ↳ Luma does this in a pretty unique way, the controls are activated with your prompt—type “camera” and hit spacebar to bring up a drop down menu where you can choose from multiple camera control options: Pull (out) Push (in) Pan (left/right) Move (left/right) Orbit (left/right) Crane (up/down) Static 💾 Dream Machine API   ↳ Integration is available through official API and third-party services like PiAPI, it’s designed to be intuitive and user-friendly and don’t quote me but I think there plans to incorporate audio and 3D elements in future versions. Also, it has ComfyUI integration through custom nodes available for direct API calls with Comfy! And, YES! Luma Dream Machine has a Free account with 30 free monthly generation credits, enough to experiment and check out the new features in their latest model release (Dream Machine 1.6). #lumaai #dreammachine #midjourney #capcut

16

3 Comments