Loading video...

Video Failed to Load

Go Home

Trying physics-driven animation with two ragdolls. Both the local VR user avatar and the NPC avatar (which represents another user) are physically simulated This is just a start, still a lot of work to do

vr_hai's profile picture

Haï~

13,125 subscribers

32,460 views • 1 year ago •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

7 Comments

Spiritmarsrover's profile picture
Spiritmarsrover1 year ago

Nice. There is a simple interaction that I've thought about but is never simulated in vr, pushing a waist height drawer closed with your waist. I'd want to see if that works.

MayflowerZero's profile picture
MayflowerZero1 year ago

this looks amaying already, i am so curious to see where this goes! Also can i just say that you are one of the most creative persons i see on my "x" feed? :>

Konshu's profile picture
Konshu1 year ago

silly but cool

Haru Rose's profile picture
Haru Rose1 year ago

What.. what are u doing over there? ... damn that looks fun

Tylyn's profile picture
Tylyn1 year ago

..................woah

Killswig's profile picture
Killswig1 year ago

my one thought about it seeing this, if the player viewport gets moved because of the physics, thats instant nausea for me, main reason i can't play boneworks, if my hands touch something i get lifted up and bam I gotta abort vr

Haï~'s profile picture
Haï~1 year ago

I am not moving the viewport, that's not the experience I'm going for. The plan is for the local avatar to look different in first-person than in cameras; so the head of your own avatar would move in video captures if you get pushed, while it remains immovable in first-person.

Related Videos

This video shows a conversation between an AI-powered NPC character and an alien avatar user (me) in VR. The AI can answer follow-up questions based on a specified personality profile. Built entirely with Unity #vr #ar #GPT3 #unity3d
1:34

This video shows a conversation between an AI-powered NPC character and an alien avatar user (me) in VR. The AI can answer follow-up questions based on a specified personality profile. Built entirely with Unity #vr #ar #GPT3 #unity3d

Ali

13,986 views • 3 years ago

Challenge accepted. For this I attempted to bloat the avatar to simulate average VRChat user and stay on Medium at the same time. The only problem is, I suck at adding bloat. Avatar stats in the reply.
0:33

Challenge accepted. For this I attempted to bloat the avatar to simulate average VRChat user and stay on Medium at the same time. The only problem is, I suck at adding bloat. Avatar stats in the reply.

MaraRizer

16,872 views • 1 month ago

From live video to 3D avatar in seconds. This #SIGGRAPH2025 paper from Adobe Research reconstructs a realistic head avatar on the fly, with no pre-cached data, and adapts seamlessly to facial motion for VR, animation, and online communication. 🔗
0:23

From live video to 3D avatar in seconds. This #SIGGRAPH2025 paper from Adobe Research reconstructs a realistic head avatar on the fly, with no pre-cached data, and adapts seamlessly to facial motion for VR, animation, and online communication. 🔗

Adobe Research

29,019 views • 9 months ago

I started looking into physics-based animation, but I am massively struggling. (For social VR, it would be mixing the solved avatar IK pose with a physics simulation, maybe with an additional specialized IK pass to keep the head joined with the HMD)
0:41

I started looking into physics-based animation, but I am massively struggling. (For social VR, it would be mixing the solved avatar IK pose with a physics simulation, maybe with an additional specialized IK pass to keep the head joined with the HMD)

Haï~

19,185 views • 1 year ago

Bring your drawing and 2D style to the VR World Having Your own avatar is easier than ever. 2D avatar, your new way to create avatars with just drawings here on #VRChat #VR #3d #shader
0:31

Bring your drawing and 2D style to the VR World Having Your own avatar is easier than ever. 2D avatar, your new way to create avatars with just drawings here on #VRChat #VR #3d #shader

DX

123,814 views • 8 months ago

With the scientists at Mikoverse we are doing research with jiggle physics. Here we see a scientist playing with the jiggle of another user. There are some latency that we are working through, but the potential is there. This is for education and scientific purposes only.
0:20

With the scientists at Mikoverse we are doing research with jiggle physics. Here we see a scientist playing with the jiggle of another user. There are some latency that we are working through, but the potential is there. This is for education and scientific purposes only.

CodeMiko

147,408 views • 2 years ago

It looks like this in VR. The point would be for the VR avatars of players to be controlled by combining both physics animation and a multi-pass IK solver. Physics animation alone won't be good (disjointed body feeling, loss of proprioception and control).
0:38

It looks like this in VR. The point would be for the VR avatars of players to be controlled by combining both physics animation and a multi-pass IK solver. Physics animation alone won't be good (disjointed body feeling, loss of proprioception and control).

Haï~

21,886 views • 1 year ago

I'm honored to share this peek into the production design of Avatar: Fire and Ash. This represents the work of so many talented teams between LA and NZ.
5:31

I'm honored to share this peek into the production design of Avatar: Fire and Ash. This represents the work of so many talented teams between LA and NZ.

Dylan Cole Art

281,183 views • 4 months ago

New Puplic Lycanroc Drone Avatar A new public SDL drone avatar has been released. A Lycanroc drone avatar. The avatar has the following options: - Colour can be changed - Hypnovisor with effect Textured and modified by Roc Unit L-049 . We hope you all like this new avatar and have fun with it. The avatar is Adviable on our SDL avatar World. Links to it in the comments. A quest version of the avatar will be released in the coming days.
1:07

New Puplic Lycanroc Drone Avatar A new public SDL drone avatar has been released. A Lycanroc drone avatar. The avatar has the following options: - Colour can be changed - Hypnovisor with effect Textured and modified by Roc Unit L-049 . We hope you all like this new avatar and have fun with it. The avatar is Adviable on our SDL avatar World. Links to it in the comments. A quest version of the avatar will be released in the coming days.

SDL

18,453 views • 5 months ago

Was doing an experiment with my HTC Cosmos Elite and decided to do a random video of the current progress of the squish dragon avatar. The avatar is far from being finished.
0:23

Was doing an experiment with my HTC Cosmos Elite and decided to do a random video of the current progress of the squish dragon avatar. The avatar is far from being finished.

Irix

82,215 views • 2 years ago

Avatar: Frontiers of Pandora has more than quadrupled its peak player count on Steam more than two years after release. Looks like a lot of new players are jumping in with the release of Avatar: Fire and Ash. A new third person mode was also added earlier this month.
0:25

Avatar: Frontiers of Pandora has more than quadrupled its peak player count on Steam more than two years after release. Looks like a lot of new players are jumping in with the release of Avatar: Fire and Ash. A new third person mode was also added earlier this month.

KAMI

1,711,612 views • 5 months ago

The physics-based animation no longer explodes every time there's a too large force colliding with the avatar rig, so compared to yesterday, it's progress 💥
0:27

The physics-based animation no longer explodes every time there's a too large force colliding with the avatar rig, so compared to yesterday, it's progress 💥

Haï~

27,857 views • 1 year ago

I must admit, this is incredibly fun and satisfying to play around with the physics There's still a lot of work in progress in this scene, but things are coming together! #screenshotsaturday #ue #gamedev
0:34

I must admit, this is incredibly fun and satisfying to play around with the physics There's still a lot of work in progress in this scene, but things are coming together! #screenshotsaturday #ue #gamedev

Spinker (Nicolas)

70,887 views • 1 year ago

James Cameron spent $400m on Avatar 3. That’s on top of $350m for Avatar 2 and $250m for the first Avatar. So, the official tally is $1B to make the entire Avatar series (but still only $17 for the logo).
2:44

James Cameron spent $400m on Avatar 3. That’s on top of $350m for Avatar 2 and $250m for the first Avatar. So, the official tally is $1B to make the entire Avatar series (but still only $17 for the logo).

Trung Phan

1,442,394 views • 5 months ago

Hey! I'm open for avatar re-texture work for VRChat! Want an avatar you will love and be proud to wear? Then I'm the fox to help you with that! Quality work and service. Accepting any avatar models! Fill this out to get a quote:
0:54

Hey! I'm open for avatar re-texture work for VRChat! Want an avatar you will love and be proud to wear? Then I'm the fox to help you with that! Quality work and service. Accepting any avatar models! Fill this out to get a quote:

Alpha WereFox

16,471 views • 9 months ago

🎉 Big news: The most iconic CEO in history just became a Synthesia user! He leads one of the world’s most recognized enterprises — one that delivers to every country in the world. And now, he’s letting YOU create free videos with his personal avatar:
0:45

🎉 Big news: The most iconic CEO in history just became a Synthesia user! He leads one of the world’s most recognized enterprises — one that delivers to every country in the world. And now, he’s letting YOU create free videos with his personal avatar:

Synthesia 🎥

88,905 views • 1 year ago

Introducing Random Avatar Matches and Avatar Arcade Mode! Mix and match special moves to create a one-of-a-kind avatar with a fighting style of your very own and take them into these two chaotic modes to earn rewards and level up. See more details via the blog:
2:07

Introducing Random Avatar Matches and Avatar Arcade Mode! Mix and match special moves to create a one-of-a-kind avatar with a fighting style of your very own and take them into these two chaotic modes to earn rewards and level up. See more details via the blog:

Street Fighter 6

1,330,950 views • 22 days ago

Want to create an avatar from a single image? FlexAvatar is a transformer model that creates full 360°, high-quality, and expressive 3D head avatar from just a single portrait image in minutes. Real-time Demo: FlexAvatar's lightweight architecture allows both animation and rendering in real-time, enabling interactive user experiences. To create a new 3D head avatar, only one image is required, e.g., from a webcam. The final avatar is ready after 2 minutes. Architecture: Under the hood, FlexAvatar adopts a transformer-based encoder-decoder design. The encoder maps the input image onto a latent avatar space, while the decoder produces 3D Gaussian attribute maps by incorporating the animation signal via cross-attention. The model learns all facial animations directly from the data without relying on pre-built 3D face models. This equips the avatars with realistic facial expressions. The internal avatar latent space can be conveniently used to integrate additional observations of a person via fitting. This enables use-cases where more than one image of a person is available, e.g., from a phone scan of the person. We train jointly on 2D monocular videos and multi-view data. However, in monocular videos, the animation signal leaks the target viewpoint, causing the model to produce incomplete 3D heads. We call this phenomenon entanglement of driving signal and target viewpoint. To prevent entanglement, we introduce bias sinks. These are learnable tokens that indicate whether a training sample stems from a monocular or a multi-view dataset. During training, the model learns to produce incomplete 3D heads only when the monocular token is present. During inference, FlexAvatar then always uses the multi-view token for which the model has learned to produce complete 3D heads. This simple design allows to combine the generalizability from monocular data with the quality of multi-view data. FlexAvatar summary: - Input: Single-image, phone scan, or monocular video - Output: Full 360° head avatar - Expressive animations - Real-time rendering and animation - Generalization to any portrait - Create a new avatar in 2 minutes - Use bias sinks to combine 2D and 3D data 🏠 🌍 🎥 Great work by Tobias Kirschstein and Simon Giebenhain!
0:30

Want to create an avatar from a single image? FlexAvatar is a transformer model that creates full 360°, high-quality, and expressive 3D head avatar from just a single portrait image in minutes. Real-time Demo: FlexAvatar's lightweight architecture allows both animation and rendering in real-time, enabling interactive user experiences. To create a new 3D head avatar, only one image is required, e.g., from a webcam. The final avatar is ready after 2 minutes. Architecture: Under the hood, FlexAvatar adopts a transformer-based encoder-decoder design. The encoder maps the input image onto a latent avatar space, while the decoder produces 3D Gaussian attribute maps by incorporating the animation signal via cross-attention. The model learns all facial animations directly from the data without relying on pre-built 3D face models. This equips the avatars with realistic facial expressions. The internal avatar latent space can be conveniently used to integrate additional observations of a person via fitting. This enables use-cases where more than one image of a person is available, e.g., from a phone scan of the person. We train jointly on 2D monocular videos and multi-view data. However, in monocular videos, the animation signal leaks the target viewpoint, causing the model to produce incomplete 3D heads. We call this phenomenon entanglement of driving signal and target viewpoint. To prevent entanglement, we introduce bias sinks. These are learnable tokens that indicate whether a training sample stems from a monocular or a multi-view dataset. During training, the model learns to produce incomplete 3D heads only when the monocular token is present. During inference, FlexAvatar then always uses the multi-view token for which the model has learned to produce complete 3D heads. This simple design allows to combine the generalizability from monocular data with the quality of multi-view data. FlexAvatar summary: - Input: Single-image, phone scan, or monocular video - Output: Full 360° head avatar - Expressive animations - Real-time rendering and animation - Generalization to any portrait - Create a new avatar in 2 minutes - Use bias sinks to combine 2D and 3D data 🏠 🌍 🎥 Great work by Tobias Kirschstein and Simon Giebenhain!

Matthias Niessner

76,454 views • 5 months ago

His hatred for the protagonist is the reason a lot of people watch Avatar
0:53

His hatred for the protagonist is the reason a lot of people watch Avatar

Unique Movie Moments 🐬

598,943 views • 5 months ago

The Avatar scandal just went from bad to worse This is painful, and one of the greatest tragedies in animation history Overnight loss of control of over a decade of IP growth I am obsessed with the Avatar animation franchise & this is just painful to see. Because, based on what I have seen with these clips, the movie would have been mind-blowing! Marketing is messed up They've lost control of their IP, and also, I don't even know how they will recover from this If you are given the chance to watch the full movie, will you watch it now or still wait till October? Oh, and by the way, Aang >>> Korra.
0:36

The Avatar scandal just went from bad to worse This is painful, and one of the greatest tragedies in animation history Overnight loss of control of over a decade of IP growth I am obsessed with the Avatar animation franchise & this is just painful to see. Because, based on what I have seen with these clips, the movie would have been mind-blowing! Marketing is messed up They've lost control of their IP, and also, I don't even know how they will recover from this If you are given the chance to watch the full movie, will you watch it now or still wait till October? Oh, and by the way, Aang >>> Korra.

Debayo 🌱

1,146,395 views • 1 month ago