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.

Another beast from Alibaba. Live Avatar: real-time, infinite-length streaming avatar generator; - 20 FPS, based on a 14B WanS2V trained with DMD. - locks identity; They plan to: add TTS, low VRAM, 3 steps, SVD quantz...

Wildminder

9,519 subscribers

21,313 views • 8 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

This is wild. China's Alibaba just dropped Live Avatar. This AI turns any voice into a realtime, talking avatar with infinite length at 20 FPS. 10 wild demos:👇 1. Ilya interview that never happened

Min Choi

220,784 views • 8 months ago

Xmax X1, the first real-time interactive video model, is here. Powered by autoregressive streaming generation, X1 achieves millisecond ultra-low latency and infinite-length generation, enabling truly natural spatial interaction. Camera is redefined. It’s no longer just a lens, but a magic wand that breaks the barrier between dimensions. Summon virtual beings into your reality and interact with them in real-time. Live implementation video is below, truely amazing work! Definitely one to watch.

Xmax X1, the first real-time interactive video model, is here. Powered by autoregressive streaming generation, X1 achieves millisecond ultra-low latency and infinite-length generation, enabling truly natural spatial interaction. Camera is redefined. It’s no longer just a lens, but a magic wand that breaks the barrier between dimensions. Summon virtual beings into your reality and interact with them in real-time. Live implementation video is below, truely amazing work! Definitely one to watch.

机器之心 JIQIZHIXIN

91,853 views • 6 months ago

Bumble CEO Whitney Wolfe Herd announces the future of dating will be decided by AI One person's AI avatar will get to know another person's AI avatar until they decide that the real-life people should meet each other Yes, this is real AI-powered “dating concierges” are planned to be the h future of these apps - Users create a personal AI avatar concierge - It will be based on your profile, values, insecurities, communication style - That AI then interacts with hundreds or thousands of other users’ AIs “going on dates,” chatting, assessing compatibility - The AI narrows it down to the best 3–5 real humans for you to meet in person This takes the human connection completely out of the equation and should be avoided at all costs Black Mirror was a documentary But I found actually gets worse Bumble’s CEO is also pushing a major Bumble redesign in late 2026. This will kill the classic swipe feature entirely in favor of AI-driven matching

Wall Street Apes

51,430 views • 3 months ago

I heard VRC was buggy, but this is RIDICULOUS Finally finished my low poly VRC avatar based on Adfree.EXE 👾⚠️【SVA】's design . Been working on this over the last few months off and on to study low poly modeling, and I can finally call it complete Everything is made from the ground up. Modeling, texture work (though referenced from Adfrees ref sheet) weight painting, and naturally I gave it some face tracking. I also finally had an excuse to use Poiyomi Pro's vertex glitch and vertex rounding features :D This was a super fun project and it feels great to see it finally completed C: Maybe next time I'll go for making an avatar with a higher poly count and more detail (Before anyone asks, I don't plan to make this clonable or share it since I don't own the design lol)

Threevee

46,903 views • 6 months 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!

Matthias Niessner

96,186 views • 7 months ago

I made a digital twin of myself from 10 seconds of video. In the clip: left is the real me, middle is a leading avatar model, right is Mirage Avatar X. Watch the eyes. The difference is not subtle. I have been testing AI avatar models since my first clone in 2023. Every one of them was impressive for about 30 seconds, then your brain caught up. Still eyes. One polite expression. A mouth doing all the work. Avatar X is the first model where that moment never came. Here is what makes it different: It is trained on you. Avatar X preserves your identity. Most avatar models can copy your appearance. Avatar X captures the subtle details that make you you. The way you move, the way you express yourself, and the way you naturally deliver speech. It looks like you. It moves like you. It sounds like you. It understands non-verbal performance Laughing, crying, yawning, sighing. These are the moments where most avatar models fall apart, trying to lip-sync through sounds that aren't words. Avatar X responds naturally, generating realistic facial expressions and micro-expressions instead of forcing every sound into speech. The expression goes beyond the lips Expressions are driven by the audio, through the whole face and body. Ask a question and it furrows its brows and shrugs on the tone. No other model does this to this degree. No quality degradation The first second and the last second look the same. Other models lose quality the longer the video runs. 10 seconds of input That is the entire requirement. Other models need 15 seconds, some even 1 to five minutes. Three years ago my AI clone was a party trick. This one can carry my face, my expressions and my delivery without me in the room. The bar for AI avatars just moved. Avatar X is live today. → Try it here:

Linus ✦ Ekenstam

20,752 views • 17 days ago

Meet LongCat-Video-Avatar: a robust audio-driven avatar model that pushes the boundaries of long-form video generation. Compared with the previous InfiniteTalk, LongCat-Video-Avatar delivers far better long-sequence stability and realism. New highlights： ⚙ Built on the LongCat-Video architecture, now supporting Audio-Text-to-Video (AT2V), Audio-Text-Image-to-Video (ATI2V), and Video Continuation modes. 🎭 Open-source SOTA Realism: Ranked 1st in overall anthropomorphism scores for both single and multi-subject scenarios in EvalTalker evaluations (492 participants, 3 independent raters per video). ♾ High-quality long videos: Cross-Chunk Latent Stitching prevents pixel degradation and error accumulation over time, ensuring seamless stitching quality. 🔒 Long-term consistency: Reference Skip Attention maintains ID consistency while eliminating rigid copy-paste artifacts. 🪄 Supports multi-person and infinite-length video generation. 🔗Open-sourced Code: Hugging Face: Project: Paper:

Meituan LongCat

28,170 views • 8 months ago

Introducing MIRA. A playable, multiplayer world model. A dream of Rocket League. Trained on 10k hours of data collected with publicly available bots, MIRA learns the dynamics of a four-player game. The model runs in real time at 20 fps, based on the keys you and the other players press. Built by General Intuition and kyutai, in collaboration with Epic Games. Not used to develop Rocket League. ▶️ Play the demo, read the technical report, and explore the open-source code at At ICML? Find us at Booth 111 to try it yourself and dig into the results with the team.

Introducing MIRA. A playable, multiplayer world model. A dream of Rocket League. Trained on 10k hours of data collected with publicly available bots, MIRA learns the dynamics of a four-player game. The model runs in real time at 20 fps, based on the keys you and the other players press. Built by General Intuition and kyutai, in collaboration with Epic Games. Not used to develop Rocket League. ▶️ Play the demo, read the technical report, and explore the open-source code at At ICML? Find us at Booth 111 to try it yourself and dig into the results with the team.

General Intuition

497,059 views • 1 month ago

Homeland Avatar Mode is LIVE! Lunacians, it’s time! Step into a whole new dimension of Homeland with Avatar Mode. Roam free in Lunacia like never before. No land? No problem! Your axie is your all-access pass to this feature. 🌄 Visit land plots or gather in Lunacia’s villages 🤝 Chat with friends and family or just hangout with fellow Lunacians 🎉 Join the launch celebration live on Twitch with Jihoz, Gila, and Phin today at 2:30 PM UTC. 🐰 Starting 3 PM UTC today, take part in the Wild Noctid Chase event and win gifts from our new friends. 🎍 Join the Plot Party event starting now and invite as many Lunacians as you can to visit your plot. Full announcement 👇🏼 📜 :

Axie Infinity

734,143 views • 2 years ago

Today we’re introducing a new plan for teams. HeyGen for Business. In 2025, our team interviewed hundreds of businesses using HeyGen to understand what were the roadblocks. What slowed them down were the small limits that add up once multiple departments were using HeyGen. One department hits credit caps. Another waits for avatar slots. A third can't integrate with their LMS. So we rebuilt the plan to give teams more room to create and access to features that were previously gated behind our enterprise plans. What’s included: → 5× more generation credits → 5 custom avatar slots for multiple team members → Self-serve SSO for easy, secure access → Interactive video using quizzes, branching scenarios, & embedded links for education → Screen recorder for screen capture in Studio → $20 per additional seats → Translate videos up to 60 min length → Priority processing for faster renders when you need them → SCORM export to plug courses directly into your LMS → Integrations with Zapier, Make, N8N, and HubSpot → Proofread and edit capabilities for translated videos HeyGen for Business is built for teams that are already moving.

HeyGen

222,762 views • 7 months ago

GS^3: Efficient Relighting with Triple Gaussian Splatting Abstract: We present a spatial and angular Gaussian based representation and a triple splatting process, for real-time, high-quality novel lighting-and-view synthesis from multi-view point-lit input images. To describe complex ap pearance, we employ a Lambertian plus a mixture of angular Gaussians as an effective reflectance function for each spatial Gaussian. To generate self-shadow, we splat all spatial Gaussians towards the light source to obtain shadow values, which are further refined by a small multi-layer perceptron. To compensate for other effects like global illumination, another network is trained to compute and add a per-spatial-Gaussian RGB tuple. The effectiveness of our representation is demonstrated on 30 samples with a wide variation in geometry (from solid to fluffy) and appearance (from translucent to anisotropic), as well as using different forms of input data, including rendered images of synthetic/reconstructed objects, photographs captured with a handheld camera and a flash, or from a professional lightstage. We achieve a training time of 40-70 minutes and a rendering speed of 90 fps on a single commodity GPU. Our results compare favorably with state-of-the-art techniques in terms of quality/performance.

GS^3: Efficient Relighting with Triple Gaussian Splatting Abstract: We present a spatial and angular Gaussian based representation and a triple splatting process, for real-time, high-quality novel lighting-and-view synthesis from multi-view point-lit input images. To describe complex ap pearance, we employ a Lambertian plus a mixture of angular Gaussians as an effective reflectance function for each spatial Gaussian. To generate self-shadow, we splat all spatial Gaussians towards the light source to obtain shadow values, which are further refined by a small multi-layer perceptron. To compensate for other effects like global illumination, another network is trained to compute and add a per-spatial-Gaussian RGB tuple. The effectiveness of our representation is demonstrated on 30 samples with a wide variation in geometry (from solid to fluffy) and appearance (from translucent to anisotropic), as well as using different forms of input data, including rendered images of synthetic/reconstructed objects, photographs captured with a handheld camera and a flash, or from a professional lightstage. We achieve a training time of 40-70 minutes and a rendering speed of 90 fps on a single commodity GPU. Our results compare favorably with state-of-the-art techniques in terms of quality/performance.

MrNeRF

17,786 views • 1 year 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

$I nearly 2x'd the speed while only using +1GB VRAM with the new MTP update in llama.cpp 🤯 You need to add these flags to start using it: --spec-type draft-mtp \ --spec-draft-p-min 0.75 \ --spec-draft-n-max 2 My results with Qwen3.6 27B on a single RTX 5080 ↓ ⚪️ no flag (without mtp) → 54.3 tok/s with 13.26GB VRAM 🔵 --spec-draft-n-max 2 → 90.7 tok/s with 14.29GB VRAM 🔴 --spec-draft-n-max 2 --spec-draft-p-min 0.75 → 93.9 tok/s with 14.30GB VRAM 🟢 --spec-draft-n-max 6 --spec-draft-p-min 0.75 → 93.9 tok/s with 14.87GB VRAM Increasing to 6 draft tokens didn't help my setup for some reason. I made sure to test with a low context length to have enough headroom and eliminate risk of vram stress. From my understanding: 1) The speed gains are very task-dependent. You need to test across a wide range of tasks to get a realistic idea of the benefits 2) We’re already running heavily quantized GGUF models (Q3, Q4, Q6, etc.), so we already benefit from strong speed/performance thanks to the reduced size. That’s why some people are seeing little to no improvement compared to MLX or other quantized versions The progress over the past few days has been insane to say the least. However, MTP now consumes significantly more VRAM. Personally 16GB just isn't enough to use MTP and run it with a good context size. Time to upgrade lads, 24GB+ users are eating GOOD today 🔥 Full setup below ↓$

I nearly 2x'd the speed while only using +1GB VRAM with the new MTP update in llama.cpp 🤯 You need to add these flags to start using it: --spec-type draft-mtp \ --spec-draft-p-min 0.75 \ --spec-draft-n-max 2 My results with Qwen3.6 27B on a single RTX 5080 ↓ ⚪️ no flag (without mtp) → 54.3 tok/s with 13.26GB VRAM 🔵 --spec-draft-n-max 2 → 90.7 tok/s with 14.29GB VRAM 🔴 --spec-draft-n-max 2 --spec-draft-p-min 0.75 → 93.9 tok/s with 14.30GB VRAM 🟢 --spec-draft-n-max 6 --spec-draft-p-min 0.75 → 93.9 tok/s with 14.87GB VRAM Increasing to 6 draft tokens didn't help my setup for some reason. I made sure to test with a low context length to have enough headroom and eliminate risk of vram stress. From my understanding: 1) The speed gains are very task-dependent. You need to test across a wide range of tasks to get a realistic idea of the benefits 2) We’re already running heavily quantized GGUF models (Q3, Q4, Q6, etc.), so we already benefit from strong speed/performance thanks to the reduced size. That’s why some people are seeing little to no improvement compared to MLX or other quantized versions The progress over the past few days has been insane to say the least. However, MTP now consumes significantly more VRAM. Personally 16GB just isn't enough to use MTP and run it with a good context size. Time to upgrade lads, 24GB+ users are eating GOOD today 🔥 Full setup below ↓

left curve dev

30,086 views • 3 months ago

🌈 Emergibles 🌈 is live at With emerge, we learned something loud and clear: people love personalizing and remixing their digital identity. So we built the next evolution around that. Emergibles is the easiest way to launch your own personalized avatar collection, all on-chain, powered by AI. Drop in reference images, write a short prompt, and launch. When someone collects, they bring their own digital identity to the table. Every piece is personalized. Collecting is participating and marking that you were part of a moment. Putting it on-chain unlocks what you can't do anywhere else. Creators can play with scarcity, monetize directly, and build real communities around their collections. Communities self-organize and grow however they see fit. Built on and funded by the Arbitrum ecosystem. Generate abundance.

atown

13,229 views • 3 months ago

Give your voice agents a face. Introducing Lemon Slice-2 – the world's first interactive talking AI video model Think of it as a face layer for your voice agents, available today in production as an API or an embeddable widget. What makes Lemon Slice-2 great: → Easy: All you need is 1 photo → Expressive: hands and full body motion → Any character and style: if it has a face, Lemon Slice-2 can animate it Under the hood: we trained a custom, 20B-parameter video diffusion transformer, streaming at 20fps on a single GPU. Infinite-length video generation with no error accumulation. Talk to the demo avatars on our homepage. Or, add your own interactive avatar to your website. Companies and creators are already building video agents that are: → Customer support agents who screen-share and show you where to click → Lead qualification agents that replace forms with real conversations → AI personal shoppers who try on clothes and guide purchases People don't engage with chat bubbles. They engage with faces. And now you can have a 24/7 spokesperson that’s the face of your brand. We’ve partnered with , Daily, and Modal to make this happen. As part of this launch, we’re giving away a guide: “How to Add a High-Converting Video Agent to Your Website in under 10 Minutes” Retweet and comment "LEMON SLICE" below and I'll send it over 🍋

Lina Colucci

1,163,183 views • 8 months ago

As a professional gamer, Ahmad Alshrari regularly livestreams to engage with his audiences in real-time, whilst gaming competitively. Check out the video from #LEAP25, where he experiences the power of low latency live-streaming on the JACO platform whilst playing #PUBG, thanks to Huawei Cloud's advanced Cloud infrastructure in Saudi Arabia. Find Out more about our Media & Entertainment industry solutions here: #HuaweiCloud #CloudComputing #AI #LEAP2025 #Livestreaming #Gaming #Esports #LowLatency #CDN JACO جاكو

As a professional gamer, Ahmad Alshrari regularly livestreams to engage with his audiences in real-time, whilst gaming competitively. Check out the video from #LEAP25, where he experiences the power of low latency live-streaming on the JACO platform whilst playing #PUBG, thanks to Huawei Cloud's advanced Cloud infrastructure in Saudi Arabia. Find Out more about our Media & Entertainment industry solutions here: #HuaweiCloud #CloudComputing #AI #LEAP2025 #Livestreaming #Gaming #Esports #LowLatency #CDN JACO جاكو

Huawei Cloud Middle East

119,081 views • 1 year ago

Every Mega Hero is a personalized NFT, paired with a live trading bot fully configured by the player. The minting process unfolds in 4 steps: 1 - Spirit defines risk appetite: • Zen (low risk) • Dynamic (balanced) • Degen (max chaos) 2 - Weapons : up to 5 trading pairs selected (trading arsenal) 3 - Strategies : up to five strategies chosen from a curated pool of quant models, optimized for various market conditions 4 - Sparks (behavioral traits): • Resilience (hedging) • Vigilance (reaction speed) • Time Travel (historical pattern analysis) Once minted, the Hero runs a $10K virtual portfolio in real-time. At any point, it can be linked to a real trading account mirroring its virtual trades in the real world. A new class of trading companion is coming. Mega Heroes landing soon on MegaETH.

Every Mega Hero is a personalized NFT, paired with a live trading bot fully configured by the player. The minting process unfolds in 4 steps: 1 - Spirit defines risk appetite: • Zen (low risk) • Dynamic (balanced) • Degen (max chaos) 2 - Weapons : up to 5 trading pairs selected (trading arsenal) 3 - Strategies : up to five strategies chosen from a curated pool of quant models, optimized for various market conditions 4 - Sparks (behavioral traits): • Resilience (hedging) • Vigilance (reaction speed) • Time Travel (historical pattern analysis) Once minted, the Hero runs a $10K virtual portfolio in real-time. At any point, it can be linked to a real trading account mirroring its virtual trades in the real world. A new class of trading companion is coming. Mega Heroes landing soon on MegaETH.

MegaHeroes | ∑:

11,361 views • 1 year ago

SqueezeMe: Efficient Gaussian Avatars for VR TL;DR: Three of these Gaussian Splatting avatars can be run at 72 frames per second. It runs locally on a Meta Quest 3 VR headset. Abstract (excerpt): While previous methods require a desktop GPU for real-time inference of a single avatar, we aim to squeeze multiple Gaussian avatars onto a portable virtual reality headset with real-time drivable inference. We begin by training a previous work, Animatable Gaussians, on a high-quality dataset captured with 512 cameras. The Gaussians are animated by controlling a base set of Gaussians with linear blend skinning (LBS) motion, and then further adjusting them with a neural network decoder to correct their appearance. When deploying the model on a Meta Quest 3 VR headset, we find two major computational bottlenecks: the decoder and the rendering. To accelerate the decoder, we train the Gaussians in UV-space instead of pixel-space and distill the decoder to a single neural network layer. Further, we discover that neighborhoods of Gaussians can share a single corrective from the decoder, providing an additional speedup. To accelerate the rendering, we develop a custom pipeline in Vulkan that runs on the mobile GPU. Putting it all together, we run 3 Gaussian avatars concurrently at 72 FPS on a VR headset.

MrNeRF

27,104 views • 1 year ago

ALERT: Police use a drone to catch suspects in the act of trying to steal a car from a mall parking lot in Dearborn, Michigan. The suspects were in a stolen Kia when they tried stealing another car, which was captured on camera. A Flock License Plate Reader caught the plate of a stolen Kia, and within seconds, the department's Drone First Responder unit deployed a Skydio police drone. At the same time, they used access to a live feed from their Real Time Information Center. Upon the vehicle pulling into a parking spot in the Foreman Mills parking lot, one of the suspects is seen trying to break into another vehicle as police arrive. After a short chase on foot, the officers apprehended the suspects with no reported injuries. Police say the technology is helping identify crimes in real-time faster and more safely. 📹: dearborn | IG

ALERT: Police use a drone to catch suspects in the act of trying to steal a car from a mall parking lot in Dearborn, Michigan. The suspects were in a stolen Kia when they tried stealing another car, which was captured on camera. A Flock License Plate Reader caught the plate of a stolen Kia, and within seconds, the department's Drone First Responder unit deployed a Skydio police drone. At the same time, they used access to a live feed from their Real Time Information Center. Upon the vehicle pulling into a parking spot in the Foreman Mills parking lot, one of the suspects is seen trying to break into another vehicle as police arrive. After a short chase on foot, the officers apprehended the suspects with no reported injuries. Police say the technology is helping identify crimes in real-time faster and more safely. 📹: dearborn | IG

E X X ➠A L E R T S

87,778 views • 4 months ago

Mira Murati says the next AI interface is not a chatbot. It is a model that can listen, see, interrupt, and respond in real time. "The models that we work with today, they're very turn-based." "You talk, they talk, then they go off and think." "While they're thinking, it's almost like they're deaf and blind. They cannot perceive anything else about what's going on." Her new frame: interaction models. "They're not turn-based. They're more like time-based interaction." "They're continuously taking in audio, text, video, and continuously providing output." "We cut this up in chunks of 20 milliseconds... this enables you to catch things like interruptions and simultaneous speech." That is the product shift: from AI as a prompt box to AI as a high-bandwidth collaborator.

Mira Murati says the next AI interface is not a chatbot. It is a model that can listen, see, interrupt, and respond in real time. "The models that we work with today, they're very turn-based." "You talk, they talk, then they go off and think." "While they're thinking, it's almost like they're deaf and blind. They cannot perceive anything else about what's going on." Her new frame: interaction models. "They're not turn-based. They're more like time-based interaction." "They're continuously taking in audio, text, video, and continuously providing output." "We cut this up in chunks of 20 milliseconds... this enables you to catch things like interruptions and simultaneous speech." That is the product shift: from AI as a prompt box to AI as a high-bandwidth collaborator.

Karl Mehta

191,606 views • 1 month ago