Loading video...

Video Failed to Load

Go Home

If you need high-precision motion capture, check out MANUS™'s Metagloves Pro, data gloves delivering "millimeter-level precision with no occlusion." They capture every finger, joint, and micro-movement in real time:

34,996 views • 6 months ago •via X (Twitter)

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Cooking meets cinematic AI. This stunning commercial was created with Seedance 2.0 in Lart AI Delivering realistic motion, rich food details, and premium visuals that look like a real production. Create yours here 👇 Prompt: THE FLAME CHEF - 15 second high-energy cinematic cooking film, 10 rapid scenes, fast rhythmic pacing, whip-pan and speed-ramp transitions between every scene, kinetic kitchen artistry from fresh ingredients to final plated dish. Shot on ARRI Alexa 35, 35mm lens, natural film grain, rich warm color grade with deep shadows and golden highlights, professional food cinematography, photorealistic, not CGI. Dark moody restaurant kitchen, dramatic overhead lighting, steam and warm glow. SCENE 1 (0-1.5s): Chef's hand places a cast iron pan onto a glowing gas burner with confident energy, blue flame rising around the edges, quick real time then micro slow motion on the flame glow, hard whip-pan into locked close-up. SCENE 2 (1.5-3s): Rapid skilled vegetable prep, knife rocking smoothly through fresh herbs and bell peppers in rhythmic motion, hyper-fast real time with hands in controlled motion, one pepper slice spinning in brief slow motion, top-down locked shot. SCENE 3 (3-4.5s): Butter cube tossed gracefully through the air landing in the warm pan, melting instantly into golden foam and swirl, slow motion flight with speed ramp to real time on the sizzle, side macro tracking the arc. SCENE 4 (4.5-6s): Chef flips the pan, vegetables rising upward in a golden arc, warm flame glow beneath, tiny seasoning embers drifting, speed ramp from real time flip into dramatic slow motion at the peak of the toss, low hero angle. SCENE 5 (6-7.5s): Extreme macro, garlic and chili meeting hot oil, lively golden bubbles, steam rising in a soft backlit plume, 120fps slow motion with every bubble crisp, probe lens push-in through the steam. SCENE 6 (7.5-9s): Chef seasoning from height with flowing hand movement, salt crystals raining down in slow motion through a shaft of overhead light catching like snow, slow motion rain against fast hand movement, side close-up with shallow focus on the crystals. SCENE 7 (9-10.5s): Sauce poured from a steel pan in a glossy ribbon coating the dish in one continuous silky wave, steam curling, silky hypnotic slow motion, orbiting macro around the pour. SCENE 8 (10.5-12s): Fresh herbs dropped from above, leaves tumbling in slow motion landing perfectly on the glistening dish, tiny sparkle of sauce, slow motion fall with real time landing, top-down locked shot. SCENE 9 (12-13.5s): Chef wipes the plate rim in one confident swift motion and spins the plate a quarter turn, steam rising through dramatic side light, fast confident real time, close-up tracking the hand with whip-pan on the spin. SCENE 10 (13.5-15s): Final hero shot, finished dish center frame under a single overhead spotlight, steam rising in elegant curls, background fading to darkness, chef's silhouette stepping back with satisfaction, calm slow motion after the energy, slow push-in settling to a loop-friendly hold. Movement: continuous kinetic energy throughout, hands moving with professional speed and precision, constant speed ramping between hyper-fast real time and crisp slow motion for rhythm, ingredients gracefully in flight, intensity building from scene 1 to 9 then sudden calm hero ending Camera: hard whip-pans between scenes, top-down locked shots, side macro tracking, low hero angle, probe lens push-in, orbiting pour shot, close-up hand tracking, slow push-in finale, all cuts punchy and landing on beat Effects: warm flame glow, backlit steam plumes, golden oil bubbles in extreme macro, salt crystals in overhead light shafts, glossy sauce reflections, shallow depth of field, natural film grain, dark moody kitchen atmosphere, no text no watermark.

Zar⭕on

32,126 views • 1 month ago

THE DEPTH MAP TRICK THAT FIXED DANCE ACCURACY IN SEEDANCE 2.0 Feed the model a video of someone dancing and it tries to interpret everything- the person, the clothes, the lighting, the room, and somewhere in there, the movement. Feed it a depth map and there's nothing left to interpret but the motion. Most creators trying to transfer a dance to a character reference the source footage directly, then wonder why the choreography drifts. The problem isn't the model - it's that you handed it ten variables when you only wanted one. Here's the workflow 1. Lock the character reference in GPT Image 2 first -face, build, costume, so identity holds independently of whatever motion gets applied to it 2. Convert the source dance footage into a depth map instead of using the raw video -this strips out the original performer's appearance, clothing, and environment entirely 3. Feed the depth map as the motion reference and the character sheet as the identity reference- two separate inputs doing two separate jobs, not one input trying to do both 5. Let the depth map carry only spatial movement -the model receives body position and momentum with no competing information about who's moving or what they look like 6. Keep the character and motion inputs isolated throughout - the moment you mix appearance data into the motion reference, the model starts negotiating between two identities Why this works • Raw footage passes the model everything at once- performer, wardrobe, room, lighting -and the choreography competes with all of it for attention • A depth map is pure spatial information, so the only thing left to transfer is movement • Separating identity from motion means the character can stay locked while the dance stays accurate - normally you're trading one for the other • The accuracy gain isn't the model getting better, it's the model getting fewer decisions to make Use cases: ⁃ Dance and choreography transfer onto original characters ⁃ Motion capture-style workflows without motion capture ⁃ Any sequence where a specific movement needs to survive intact ⁃ Character showcase content built on existing performance footage The character sheet answers who's dancing. The depth map answers how - and keeping those two questions separate is the whole trick.

Nexlow

84,925 views • 26 days ago

🎮 𝐅𝐞𝐛𝐫𝐮𝐚𝐫𝐲 𝟐𝟎𝟐𝟓 𝐃𝐞𝐯𝐞𝐥𝐨𝐩𝐦𝐞𝐧𝐭 𝐔𝐩𝐝𝐚𝐭𝐞 🚨 Big updates across multiple areas of Varsity this month! • Coach Schemes in Dynasty Mode – Every coach now has a unique scheme and playstyle, influencing how your team plays. Do you stick with your staff and try new strategies, or hire a coach who specializes in your preferred system? • Audio & Settings Enhancements – A custom Varsity marching band drumline is being composed to add more energy to the game. Plus, menus now have sound effects, and new resolution options (16:9 and 16:10 ultrawide) have been added. • 3D Player & Stadium Models – The new player models are fully integrated, showcasing enhanced customization and animations. Pro-style stadiums for playoffs and championships are also in development! • Motion Capture Upgrades – Defensive back (DB) movements have been refined using motion capture for better realism. We’ve also improved base locomotion (running, turning, and movement) and are hiring a full-time AAA animation specialist. This month was all about refining core gameplay features and bringing more realism to Varsity. Stay tuned for more updates as we keep pushing towards launch! Support us on Patreon for Steam Pre-Alpha PlayTest access, and don’t forget to wish list on Steam! #Varsity #HSFootballVideoGame #DynastyMode #FootballGaming #HighSchoolFootball #CoachSchemes #3DModels #MotionCapture

Varsity - High School Football Video Game

99,319 views • 1 year ago

Trained on zero real-world data. Learned to walk, pick up boxes, and follow multi-step instructions... in the REAL world. ( 📌 Paper below) Researchers from Amazon FAR, Berkeley, Stanford, and CMU scanned real rooms with an iPhone, rebuilt them as 3D Gaussian Splatting scenes, then generated 48,000 synthetic trajectories of a Unitree G1 walking, grasping, and placing objects inside those virtual replicas. They rendered the robot's first-person camera view from each run and paired it with the matching language instruction and motion data. That's the dataset every humanoid team needs and nobody has: synced egocentric video + language + kinematics, at scale. Instead of collecting it in the real world, they manufactured it. They trained a vision-language-kinematics policy on that synthetic data alone, then deployed it on the physical G1 across five task types: navigation to a named object, lifting boxes of three different sizes with no per-size tuning, chained multi-step tasks, robustness to mid-task layout changes and flickering lights, and multi-minute long-horizon runs. No real-world fine-tuning at any point. Real-world interaction data has been the hard limit on humanoid learning... slow, expensive, and small. If scanning a room once and synthesizing thousands of labeled interactions holds up as a general recipe, that limit moves. Data stops being the bottleneck robotics teams have to solve for. 📌 Paper: Project: ——- Weekly robotics and AI insights. Subscribe free:

Ilir Aliu

12,950 views • 22 days ago