Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

AI retargeting without manual bone mapping. This could be big for game animation. Different skeletons and body proportions have always made reusing animation a headache. You already have the mocap, but getting it onto a new character can mean another round of bone mapping and cleanup. UMR learns how... points on body surfaces correspond, then uses that mapping to transfer motion without manually matching bones. The research focuses on humanoid robots. I see a lot of potential for game animation here. Reusing the same motion library across characters with very different proportions, with less setup for each rig, would make a big difference. I think we'll see people adapting this for 3D workflows pretty soon. For the method and experiments, here's the full paper -show more

Stefan 3D AI

8,373 subscribers

20,482 views • 6 days ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

4 Comments

Jason Sharp6 days ago

This feels like the future of game animation Grok video -> AI retargetting It's already the way to do 2d sprites

Daniel Johnsen6 days ago

The first motion is usually the easy part. If this also handles the ugly edge cases in retargeting, that could save game teams a lot of cleanup time.

Eugene T5 days ago

Cascadeur already solves the everyday retargeting job well

Xar6 days ago

How does this work?

Related Videos

The first version of the Fondant Face Proxy Normals is feature complete, so here's a closer look at how it is used (character by Joce ). First, align the armature to the head. If you want to attach to an existing rig, make sure the Root bone is exactly placed on the head bone. Adjust it to fit the face with the alignment bone layers. Use the debug material to help see what you are doing. Then use the other bones to adjust the shape. You can attach it to an existing rig by parenting the root bone to the head bone, or even merging the proxy rig into it (2nd clip.) The Normals rig can be posed on the fly for different shots. Since the setup actually uses a mesh object to control the curves (via Geometry Nodes), you can also easily make shapekeys on the proxy to correspond to expression keys, and then drive them (3rd clip.)

The first version of the Fondant Face Proxy Normals is feature complete, so here's a closer look at how it is used (character by Joce ). First, align the armature to the head. If you want to attach to an existing rig, make sure the Root bone is exactly placed on the head bone. Adjust it to fit the face with the alignment bone layers. Use the debug material to help see what you are doing. Then use the other bones to adjust the shape. You can attach it to an existing rig by parenting the root bone to the head bone, or even merging the proxy rig into it (2nd clip.) The Normals rig can be posed on the fly for different shots. Since the setup actually uses a mesh object to control the curves (via Geometry Nodes), you can also easily make shapekeys on the proxy to correspond to expression keys, and then drive them (3rd clip.)

aVersionOfReality

11,380 views • 1 year ago

Day 10. Building the game I've dreamed of for 15 years with AI. I still can't get over the feeling that I can finally animate all the creatures I've been imagining for this game. Animation was the part I was most worried about when I started. I thought I'd need an animator, especially for creatures like centaurs, scorpions and spiders. So today I went off-plan and made a Scorpid. It won't even be in the first version of the game. I just wanted to see whether I could get its animations working. I started with a good model made in Tripo P2.0, already optimized and split into logical parts. I think that preparation matters a lot. GPT-6 works with the mesh through code and calculations. A sensible number of vertices and well-organized parts make it easier to work out bone placement and skin weights. Then I made video references for the animations with SD 2.5 and repeated the same workflow. About an hour and a half with GPT-6 Astra on xhigh: five animations ready for the game, including several rounds of feedback. Another hour later, I had it running in Unity. I used to run a 3D studio. Work like this used to take weeks. Now I'm watching this thing move around in my game, and I still can't quite believe my eyes.

Day 10. Building the game I've dreamed of for 15 years with AI. I still can't get over the feeling that I can finally animate all the creatures I've been imagining for this game. Animation was the part I was most worried about when I started. I thought I'd need an animator, especially for creatures like centaurs, scorpions and spiders. So today I went off-plan and made a Scorpid. It won't even be in the first version of the game. I just wanted to see whether I could get its animations working. I started with a good model made in Tripo P2.0, already optimized and split into logical parts. I think that preparation matters a lot. GPT-6 works with the mesh through code and calculations. A sensible number of vertices and well-organized parts make it easier to work out bone placement and skin weights. Then I made video references for the animations with SD 2.5 and repeated the same workflow. About an hour and a half with GPT-6 Astra on xhigh: five animations ready for the game, including several rounds of feedback. Another hour later, I had it running in Unity. I used to run a 3D studio. Work like this used to take weeks. Now I'm watching this thing move around in my game, and I still can't quite believe my eyes.

Stefan 3D AI

308,714 views • 7 days ago

SkildAI, which is developing a foundational AI model for robots, has released a new video of their humanoid robot walking over objects and exploring environments it had never seen before. The robot is using vision & adapts in real time. SkildAI: "From raw images and joint feedback, the model directly outputs low-level motor commands. This single neural network enables humanoid robots to seamlessly walk across flat ground, climb stairs, and step over obstacles without any planning, mapping, or manual switching between behaviors. For all of our testing in this video, we deployed the robot in each new environment without any prior planning or mapping. We took off the humanoid’s shoes so you can hear it.

SkildAI, which is developing a foundational AI model for robots, has released a new video of their humanoid robot walking over objects and exploring environments it had never seen before. The robot is using vision & adapts in real time. SkildAI: "From raw images and joint feedback, the model directly outputs low-level motor commands. This single neural network enables humanoid robots to seamlessly walk across flat ground, climb stairs, and step over obstacles without any planning, mapping, or manual switching between behaviors. For all of our testing in this video, we deployed the robot in each new environment without any prior planning or mapping. We took off the humanoid’s shoes so you can hear it.

Sawyer Merritt

69,381 views • 1 year ago

Here is DLSS5 working on an animation I made in Unreal. It uses the existing geometry, animation, textures and lighting to render. It doesn't make stuff up out of thin air. There is zero generative AI. This is the exact use of AI that people have wanted. As a tool where artists maintain full art direction. Neural rendering makes creating amazing art with real 3D models easier and more accessible. All 3D artists should be excited. Here is another explanation by Daniel Vávra ⚔ Is this a 3D model? I encourage you to create a 3D model or animation sometime and try it out for yourself, it can be fun creating something!

Here is DLSS5 working on an animation I made in Unreal. It uses the existing geometry, animation, textures and lighting to render. It doesn't make stuff up out of thin air. There is zero generative AI. This is the exact use of AI that people have wanted. As a tool where artists maintain full art direction. Neural rendering makes creating amazing art with real 3D models easier and more accessible. All 3D artists should be excited. Here is another explanation by Daniel Vávra ⚔ Is this a 3D model? I encourage you to create a 3D model or animation sometime and try it out for yourself, it can be fun creating something!

Unreal Sensei

125,391 views • 18 days ago

Here are my Cleanup Animation shots for Episode 8 of @HazbinHotel 😈🐍✨🎶 I have A LOT OF FU*KING COOL SHOTS! I have a cool shot with all the babies And the big revelations of the last episode!! AN INCREDIBLE EPISODE that I'm proud to be a part of it! #HazbinHotel

Here are my Cleanup Animation shots for Episode 8 of @HazbinHotel 😈🐍✨🎶 I have A LOT OF FU*KING COOL SHOTS! I have a cool shot with all the babies And the big revelations of the last episode!! AN INCREDIBLE EPISODE that I'm proud to be a part of it! #HazbinHotel

AnimatedMau

49,866 views • 2 years ago

Physics-based Motion Retargeting from Sparse Inputs paper page: Avatars are important to create interactive and immersive experiences in virtual worlds. One challenge in animating these characters to mimic a user's motion is that commercial AR/VR products consist only of a headset and controllers, providing very limited sensor data of the user's pose. Another challenge is that an avatar might have a different skeleton structure than a human and the mapping between them is unclear. In this work we address both of these challenges. We introduce a method to retarget motions in real-time from sparse human sensor data to characters of various morphologies. Our method uses reinforcement learning to train a policy to control characters in a physics simulator. We only require human motion capture data for training, without relying on artist-generated animations for each avatar. This allows us to use large motion capture datasets to train general policies that can track unseen users from real and sparse data in real-time. We demonstrate the feasibility of our approach on three characters with different skeleton structure: a dinosaur, a mouse-like creature and a human. We show that the avatar poses often match the user surprisingly well, despite having no sensor information of the lower body available. We discuss and ablate the important components in our framework, specifically the kinematic retargeting step, the imitation, contact and action reward as well as our asymmetric actor-critic observations. We further explore the robustness of our method in a variety of settings including unbalancing, dancing and sports motions.

AK

106,527 views • 3 years ago

I was excited about Stream Deck for over a year, and finally decided to get it. I have a lot of debug commands in the game and this device should help me working with them. Already have a setup for trailer camera controls: add/change points, change speed on the fly, etc.

I was excited about Stream Deck for over a year, and finally decided to get it. I have a lot of debug commands in the game and this device should help me working with them. Already have a setup for trailer camera controls: add/change points, change speed on the fly, etc.

Pavel Valakh | Lands of Koastalia 🐏

18,607 views • 1 year ago

I vibe coded and built a sprite animation pipeline 🛠️ (Day 22 of making the engine+game) ⬇️ Watch the video if you don't wanna read the wall of text - it directly shows what I do. Shoutout to Jidé ✨ for showing me a paper on black/white combination to get alpha, it's the cleanest method yet, and to Cursor for enabling this entire journey. If you prefer the wall of text here you go: The hardest part of using general image models for 2D sprites isn’t getting a nice-looking frame, it’s getting consistent motion across a whole sprite sheet. You can fake a sheet, but frames won’t align, timing drifts, and you end up with weird artifacts. Even if you manually cut frames + interpolate, the animation often looks “off” because each frame is basically a new interpretation, not the same character evolving over time. This is especially noticeable with public API models like gpt-image-1.5 and Nano Banana. Some custom LoRAs for open models exist, but this is intended for less techy folks. My workaround: use a video model first, then post-process into a sprite sheet. Render the animation over a solid background (white/black/magenta/green), then chroma-key it out (my engine tool supports this). If the motion stays inside the silhouette, this works surprisingly well. You can do this in almost any video editing software too! The catch: keying almost always leaves an “aura” (edge spill). My best results come from interpolating the keyed animation with a clean base sprite, so you keep crisp edges and only “borrow” motion/detail where needed. If the animation extends outside the silhouette (tree branches, hair wisps, foliage), I usually skip “true sprite animation” and do it with shaders instead. Keying can’t fully remove halos there, no matter how much feathering/tuning you do. Another annoying issue: pixel corruption. AI rarely generates a perfectly flat background (pure #000000 or #FF00FF). That tiny noise breaks clean extraction and creates crawling garbage pixels around the subject. For clean base sprites (and even PBR maps), a useful trick is generating the same asset on white + black backgrounds and deriving alpha from the difference. This is basically a matte workflow: white = opaque, black = transparent. It fixes aura… but you’d need it per-frame to fix animation, which is still hard. For simple pixel art (single-digit frames), you can sometimes generate a sprite sheet, then ask the model to recreate it on black/white while preserving alignment… but it’s still manual-heavy. Honestly, at this point, for some projects it’s easier to go 3D → 2D and render clean sprites/maps directly. But I still love pushing “pure 2D” and seeing how far we can take it. Thanks for reading! Follow/bookmark/repost if interested in this kind of content!

I vibe coded and built a sprite animation pipeline 🛠️ (Day 22 of making the engine+game) ⬇️ Watch the video if you don't wanna read the wall of text - it directly shows what I do. Shoutout to Jidé ✨ for showing me a paper on black/white combination to get alpha, it's the cleanest method yet, and to Cursor for enabling this entire journey. If you prefer the wall of text here you go: The hardest part of using general image models for 2D sprites isn’t getting a nice-looking frame, it’s getting consistent motion across a whole sprite sheet. You can fake a sheet, but frames won’t align, timing drifts, and you end up with weird artifacts. Even if you manually cut frames + interpolate, the animation often looks “off” because each frame is basically a new interpretation, not the same character evolving over time. This is especially noticeable with public API models like gpt-image-1.5 and Nano Banana. Some custom LoRAs for open models exist, but this is intended for less techy folks. My workaround: use a video model first, then post-process into a sprite sheet. Render the animation over a solid background (white/black/magenta/green), then chroma-key it out (my engine tool supports this). If the motion stays inside the silhouette, this works surprisingly well. You can do this in almost any video editing software too! The catch: keying almost always leaves an “aura” (edge spill). My best results come from interpolating the keyed animation with a clean base sprite, so you keep crisp edges and only “borrow” motion/detail where needed. If the animation extends outside the silhouette (tree branches, hair wisps, foliage), I usually skip “true sprite animation” and do it with shaders instead. Keying can’t fully remove halos there, no matter how much feathering/tuning you do. Another annoying issue: pixel corruption. AI rarely generates a perfectly flat background (pure #000000 or #FF00FF). That tiny noise breaks clean extraction and creates crawling garbage pixels around the subject. For clean base sprites (and even PBR maps), a useful trick is generating the same asset on white + black backgrounds and deriving alpha from the difference. This is basically a matte workflow: white = opaque, black = transparent. It fixes aura… but you’d need it per-frame to fix animation, which is still hard. For simple pixel art (single-digit frames), you can sometimes generate a sprite sheet, then ask the model to recreate it on black/white while preserving alignment… but it’s still manual-heavy. Honestly, at this point, for some projects it’s easier to go 3D → 2D and render clean sprites/maps directly. But I still love pushing “pure 2D” and seeing how far we can take it. Thanks for reading! Follow/bookmark/repost if interested in this kind of content!

Startracker 🔺

20,181 views • 8 months ago

[Discord] [For the Full Video check my Bio or the Link below] [Personal Comment] Hello everyone!❤️Here I am again with a little preview of the naimation I ahve been working on! I think this is by far one of my most creative videos I have done, because I tried to make some new stuff with this request, not just the simplke black screen transition but actually have a scene how to change from feet throatlift to the actual throatlift and I think this was a lot of fun, you probably won't see this clip on the Preview, but if you want to see it, just check out the links! And also, have a great day everyone, I think you will still like this preview a lot because I was so nice and made it extra long for you, even though 30sec is a really long time considering that the entire video is only 2min and 20sec long which also marks a new record, for the longest animation yet! With that being said, I wish you all a beuatiful day and see you all the next days, when I release the new animation! ❤️❤️😊 [Links] If you want to see more, make requests etc.-> [Tags] #strangling #strangled #choke #choked #choking #throatlifted #ネックハンギングツリー #首絞め #request #Xianyun #Kafka #Feet

[Discord] [For the Full Video check my Bio or the Link below] [Personal Comment] Hello everyone!❤️Here I am again with a little preview of the naimation I ahve been working on! I think this is by far one of my most creative videos I have done, because I tried to make some new stuff with this request, not just the simplke black screen transition but actually have a scene how to change from feet throatlift to the actual throatlift and I think this was a lot of fun, you probably won't see this clip on the Preview, but if you want to see it, just check out the links! And also, have a great day everyone, I think you will still like this preview a lot because I was so nice and made it extra long for you, even though 30sec is a really long time considering that the entire video is only 2min and 20sec long which also marks a new record, for the longest animation yet! With that being said, I wish you all a beuatiful day and see you all the next days, when I release the new animation! ❤️❤️😊 [Links] If you want to see more, make requests etc.-> [Tags] #strangling #strangled #choke #choked #choking #throatlifted #ネックハンギングツリー #首絞め #request #Xianyun #Kafka #Feet

Leandra190

16,447 views • 2 years ago

Took Grok 4.6 out for a spin this weekend and built this horde shooter prototype on ThreeJS. It is a big step up from Grok 4.5! I've always wanted to try building a horde shooter game with thousands of enemies onscreen and used Grok 4.6 for pretty much everything (including UI, vfx, and even the sfx!), except the 3D models for the characters, which were made and rigged with Meshy. I'm also amazed that ThreeJS with WebGPU can really handle thousands of enemies onscreen at the same time, though I think I pushed the limit here with Depth of Field and some other fancy shaders. Very fun!

Took Grok 4.6 out for a spin this weekend and built this horde shooter prototype on ThreeJS. It is a big step up from Grok 4.5! I've always wanted to try building a horde shooter game with thousands of enemies onscreen and used Grok 4.6 for pretty much everything (including UI, vfx, and even the sfx!), except the 3D models for the characters, which were made and rigged with Meshy. I'm also amazed that ThreeJS with WebGPU can really handle thousands of enemies onscreen at the same time, though I think I pushed the limit here with Depth of Field and some other fancy shaders. Very fun!

Danny Limanseta

65,105 views • 1 month ago

We’ve updated our Insta360 X5 Mapping Guide. 📍 One of the most important changes is simple: keep each mapping session under 10 minutes and return to where you started. Ideally, walk for around 5 minutes, turn around, and map your way back to the same starting point. Why? Because the way you capture a location directly affects the quality of the 3D reconstruction. Better capture paths mean better geometry, more reliable reconstructions, and higher-quality Gaussian Splats like the one you see here. If you’re mapping with an Insta360 X5, make sure to follow the updated guide before your next session. Read the updated Mapping Guide: Own an Insta360 X5? Start mapping and earning with OVER: Using another 360° camera? Generate your Gaussian Splat for free on FreeGaussian, powered by OVER:

We’ve updated our Insta360 X5 Mapping Guide. 📍 One of the most important changes is simple: keep each mapping session under 10 minutes and return to where you started. Ideally, walk for around 5 minutes, turn around, and map your way back to the same starting point. Why? Because the way you capture a location directly affects the quality of the 3D reconstruction. Better capture paths mean better geometry, more reliable reconstructions, and higher-quality Gaussian Splats like the one you see here. If you’re mapping with an Insta360 X5, make sure to follow the updated guide before your next session. Read the updated Mapping Guide: Own an Insta360 X5? Start mapping and earning with OVER: Using another 360° camera? Generate your Gaussian Splat for free on FreeGaussian, powered by OVER:

Over the Reality 🌐

48,693 views • 15 days ago

I've built a setup in Geometry Nodes to get data from bones, so now I can do things like make Normals from them, and I'm experimenting with much more. The Normal here is basically a cylinder around each bone. So it doesn't work that great around the shoulders, but its great on the arms and legs. Its the same idea as when you do radial/sphere Normals for a face or hair, but for the whole body.

I've built a setup in Geometry Nodes to get data from bones, so now I can do things like make Normals from them, and I'm experimenting with much more. The Normal here is basically a cylinder around each bone. So it doesn't work that great around the shoulders, but its great on the arms and legs. Its the same idea as when you do radial/sphere Normals for a face or hair, but for the whole body.

aVersionOfReality

13,309 views • 11 months ago

Thoughts on transferable skills. I was surprised reading the comments to Riley’s question about best UI animation tool for web. To anyone interested in diving into motion design I would strongly recommend - aim to become a motion expert, not Jitter/Rive/Lottie specialist. If you start from the foundations, jumping into any animation tool will feel like a walk in the park. Keyframes, easing functions, transformations - same concepts are shared between all of them. Getting to know the foundations on the web means starting with CSS, Tailwind or React Motion. I’m often creating small motion concepts in Figma (see video) and then recreating them manually in code. Any language, it doesn’t matter because principles are the same. I recently made the switch to Swift and built a small animation tool for myself (see video) that I’ll discard after I’m done with the motion. This level of comfortableness with motion gives you unmatched flexibility. From that point, tools are just interfaces for expressing something you already deeply understand. This is not about performance or technical optimizations (Lottie files can outperform CSS equivalents), tools mentioned above really are great - it’s about developing transferable creative skill that lets you think across any medium, platform or tool.

Thoughts on transferable skills. I was surprised reading the comments to Riley’s question about best UI animation tool for web. To anyone interested in diving into motion design I would strongly recommend - aim to become a motion expert, not Jitter/Rive/Lottie specialist. If you start from the foundations, jumping into any animation tool will feel like a walk in the park. Keyframes, easing functions, transformations - same concepts are shared between all of them. Getting to know the foundations on the web means starting with CSS, Tailwind or React Motion. I’m often creating small motion concepts in Figma (see video) and then recreating them manually in code. Any language, it doesn’t matter because principles are the same. I recently made the switch to Swift and built a small animation tool for myself (see video) that I’ll discard after I’m done with the motion. This level of comfortableness with motion gives you unmatched flexibility. From that point, tools are just interfaces for expressing something you already deeply understand. This is not about performance or technical optimizations (Lottie files can outperform CSS equivalents), tools mentioned above really are great - it’s about developing transferable creative skill that lets you think across any medium, platform or tool.

Natko Hasic

64,045 views • 1 year ago

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

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

Martin Nebelong

48,280 views • 2 years ago

This Roman Real-Time Environment is a project by Junior 3D Artist Davide Maddalo for presentation in Unreal Engine 5. Modelled mainly in Maya, the project was a practice in exploring the full real-time pipeline with a focus on how all aspects, from assets to lighting, work together. "One feature I found particularly useful while working on this environment was tri-planar mapping. I learned how it can help project textures consistently across surfaces without relying entirely on traditional UV mapping, which was especially useful for creating more flexible and natural-looking materials within the environment." - Davide Maddalo

This Roman Real-Time Environment is a project by Junior 3D Artist Davide Maddalo for presentation in Unreal Engine 5. Modelled mainly in Maya, the project was a practice in exploring the full real-time pipeline with a focus on how all aspects, from assets to lighting, work together. "One feature I found particularly useful while working on this environment was tri-planar mapping. I learned how it can help project textures consistently across surfaces without relying entirely on traditional UV mapping, which was especially useful for creating more flexible and natural-looking materials within the environment." - Davide Maddalo

Unreal Engine

20,469 views • 14 days ago

SAM ALTMAN WAS JUST ASKED IF OPENAI'S HUMANOID ROBOT IS STILL ON TRACK FOR 2027 He thinks everything in bio and material science is about to be transformed, along with software that gets built on the fly for whatever you need, cybersecurity, and new kinds of energy, where he said we could see "a real renaissance." On the robot, he's sticking with 2027 for a demo but not for anything walking around. "I think we'll have a cool demo in 2027, like a great demo. I don't think that there will be like robots walking the streets for a few years longer, but we'll have something impressive."

SAM ALTMAN WAS JUST ASKED IF OPENAI'S HUMANOID ROBOT IS STILL ON TRACK FOR 2027 He thinks everything in bio and material science is about to be transformed, along with software that gets built on the fly for whatever you need, cybersecurity, and new kinds of energy, where he said we could see "a real renaissance." On the robot, he's sticking with 2027 for a demo but not for anything walking around. "I think we'll have a cool demo in 2027, like a great demo. I don't think that there will be like robots walking the streets for a few years longer, but we'll have something impressive."

IPO Newsroom

36,620 views • 6 days ago

NVIDIA MotionBricks brings humanoid robots closer to natural, fluid movement The system combines a modular latent generative model with Smart Primitives, packing more than 350,000 motion skills into a single backbone. It runs in real time at up to 15,000 FPS with just 2ms latency. Navigation, object interaction, sitting, jumping, obstacle crossing, and even different movement styles can be composed zero-shot, like reusable motion modules. It is already being used for whole-body control on the Unitree G1 and serves as the motion layer for NVIDIA GR00T Whole-Body Control. What makes this interesting is the move from training individual behaviors toward building a shared library of motion capabilities. If that approach scales, robot movement could become more modular, transferable, and easier to deploy across different tasks and environments.

NVIDIA MotionBricks brings humanoid robots closer to natural, fluid movement The system combines a modular latent generative model with Smart Primitives, packing more than 350,000 motion skills into a single backbone. It runs in real time at up to 15,000 FPS with just 2ms latency. Navigation, object interaction, sitting, jumping, obstacle crossing, and even different movement styles can be composed zero-shot, like reusable motion modules. It is already being used for whole-body control on the Unitree G1 and serves as the motion layer for NVIDIA GR00T Whole-Body Control. What makes this interesting is the move from training individual behaviors toward building a shared library of motion capabilities. If that approach scales, robot movement could become more modular, transferable, and easier to deploy across different tasks and environments.

CyberRobo

11,089 views • 3 months ago

Here is the Geometry Nodes Weighted Normals with Laplacian Blur on a full character (a vroid). It easily improves the shading even on game topology with almost no setup. I built this as part of my quest to improve real time toon shading. 3D anime models are popular, but use of dynamic light is rare even among high quality vtuber models. This is for several reasons, but a big one is simply that it takes a lot of Custom Normals work to make 3D cel shading not look like a jagged mess (other pieces of the puzzle are issues like deformations, multiple lights, etc). And fixing Normals is tedious, especially on existing game topology. I have focused on proxy meshes for priority areas like character faces, but they aren't an efficient solution for the whole body + outfit. I wanted something I could just throw on any model and make it at least not a jagged mess anymore even if it wasn't perfect. As you can see from this clip, this does that very well! And vertex groups can be used to control the style of the effect and power. It still can't smooth beyond what the topology density can support, but the topo itself is no longer a problem (for higher res, could be run on a subdivided version of the model and then baked to a Normal Map.) The only changes I made to this model were adding a weld modifier to merge split edges during interpolation, and a vertex group to select the skirt. I have not yet added full handling and logic for detecting edges with big angles like the skirt, or for handling boundaries like on the hair, so both those areas can get better too. You can also see that while it successfully smooths out the Face, it isn't really stylistically correct there. That is still best done with a proxy mesh to define a new shape. This is part of the tools I am working on for Fondant. We are putting together a Blender Addon to release this + a proxy mesh tool for the face, and are working on resolving other problems in-engine to fully bring dynamic light to real time 3D toon shading. Give us a follow, and send them a DM if you are interested in testing these tools as they develop!

Here is the Geometry Nodes Weighted Normals with Laplacian Blur on a full character (a vroid). It easily improves the shading even on game topology with almost no setup. I built this as part of my quest to improve real time toon shading. 3D anime models are popular, but use of dynamic light is rare even among high quality vtuber models. This is for several reasons, but a big one is simply that it takes a lot of Custom Normals work to make 3D cel shading not look like a jagged mess (other pieces of the puzzle are issues like deformations, multiple lights, etc). And fixing Normals is tedious, especially on existing game topology. I have focused on proxy meshes for priority areas like character faces, but they aren't an efficient solution for the whole body + outfit. I wanted something I could just throw on any model and make it at least not a jagged mess anymore even if it wasn't perfect. As you can see from this clip, this does that very well! And vertex groups can be used to control the style of the effect and power. It still can't smooth beyond what the topology density can support, but the topo itself is no longer a problem (for higher res, could be run on a subdivided version of the model and then baked to a Normal Map.) The only changes I made to this model were adding a weld modifier to merge split edges during interpolation, and a vertex group to select the skirt. I have not yet added full handling and logic for detecting edges with big angles like the skirt, or for handling boundaries like on the hair, so both those areas can get better too. You can also see that while it successfully smooths out the Face, it isn't really stylistically correct there. That is still best done with a proxy mesh to define a new shape. This is part of the tools I am working on for Fondant. We are putting together a Blender Addon to release this + a proxy mesh tool for the face, and are working on resolving other problems in-engine to fully bring dynamic light to real time 3D toon shading. Give us a follow, and send them a DM if you are interested in testing these tools as they develop!

aVersionOfReality

14,904 views • 1 year ago