正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

Splatter a Video can turn a video into a 3D Gaussian representation! This allows for enhanced video tracking, depth prediction, motion and appearance editing, and stereoscopic video generation.

Dreaming Tulpa 🥓👑

56,418 subscribers

11,586 次观看 • 2 年前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

3 条评论

Dreaming Tulpa 🥓👑 的头像

Dreaming Tulpa 🥓👑2 年前

If you like Tweets like this, you might enjoy my weekly newsletter, #aiartweekly. A free, once–weekly e-mail round-up of the latest AI art news, interviews with artists and useful tools & resources. Join 3500+ subscribers here:

Pseudonym 🦅 的头像

Pseudonym 🦅2 年前

Couple this with diffusion and I’m convinced we will have 3d replays of our videos that we can just walk around in We literally can build this now craziness

Out 的头像

Out2 年前

The fact that it doesn't use hard coded intrinsics to generates those splats it's definitely a plus point

相关视频

Introducing 3D Motion. You can now turn any image into a 3D motion video on Morphic.

Introducing 3D Motion. You can now turn any image into a 3D motion video on Morphic.

Morphic

222,108 次观看 • 1 年前

Tracking Everything Everywhere All at Once paper page: present a new test-time optimization method for estimating dense and long-range motion from a video sequence. Prior optical flow or particle video tracking algorithms typically operate within limited temporal windows, struggling to track through occlusions and maintain global consistency of estimated motion trajectories. We propose a complete and globally consistent motion representation, dubbed OmniMotion, that allows for accurate, full-length motion estimation of every pixel in a video. OmniMotion represents a video using a quasi-3D canonical volume and performs pixel-wise tracking via bijections between local and canonical space. This representation allows us to ensure global consistency, track through occlusions, and model any combination of camera and object motion. Extensive evaluations on the TAP-Vid benchmark and real-world footage show that our approach outperforms prior state-of-the-art methods by a large margin both quantitatively and qualitatively.

Tracking Everything Everywhere All at Once paper page: present a new test-time optimization method for estimating dense and long-range motion from a video sequence. Prior optical flow or particle video tracking algorithms typically operate within limited temporal windows, struggling to track through occlusions and maintain global consistency of estimated motion trajectories. We propose a complete and globally consistent motion representation, dubbed OmniMotion, that allows for accurate, full-length motion estimation of every pixel in a video. OmniMotion represents a video using a quasi-3D canonical volume and performs pixel-wise tracking via bijections between local and canonical space. This representation allows us to ensure global consistency, track through occlusions, and model any combination of camera and object motion. Extensive evaluations on the TAP-Vid benchmark and real-world footage show that our approach outperforms prior state-of-the-art methods by a large margin both quantitatively and qualitatively.

AK

280,570 次观看 • 3 年前

Gaussian Garments: Reconstructing Simulation-Ready Clothing with Photorealistic Appearance from Multi-View Video Contribution quote from the paper: In summary, our main contributions are • a comprehensive pipeline for reconstructing the shape, appearance, and behavior of real-world garments using Gaussian splatting, • an algorithm for registering garment meshes to multi- view videos with an optimization procedure based on Gaussian splatting, and • a Gaussian Garment representation that combines triangle meshes with Gaussian textures to capture photorealistic appearance and can be used as a fully controllable 3D asset.

Gaussian Garments: Reconstructing Simulation-Ready Clothing with Photorealistic Appearance from Multi-View Video Contribution quote from the paper: In summary, our main contributions are • a comprehensive pipeline for reconstructing the shape, appearance, and behavior of real-world garments using Gaussian splatting, • an algorithm for registering garment meshes to multi- view videos with an optimization procedure based on Gaussian splatting, and • a Gaussian Garment representation that combines triangle meshes with Gaussian textures to capture photorealistic appearance and can be used as a fully controllable 3D asset.

MrNeRF

27,354 次观看 • 2 年前

This is insane. This AI can take a regular 2D video clip and create a 3d version in a few seconds. This is pretty big for the future of content creation and video editing.

This is insane. This AI can take a regular 2D video clip and create a 3d version in a few seconds. This is pretty big for the future of content creation and video editing.

Barsee 🐶

286,535 次观看 • 3 年前

Gaussian Shell Maps are a new neural scene representation that connects fields and 3D Gaussians. This representation unlocks the full potential of 3D Gaussian splatting for generative AI applications, such as 3D avatar generation. 1/2

Gordon Wetzstein

52,555 次观看 • 2 年前

Introducing Claude Opus 5.5 for launch videos. Give it an idea, your assets, and a visual reference. Through the Motion MCP, it can build animated scenes, typography, and transitions into a video you can keep editing. Want something changed? Tell Claude, and it updates the video in Motion. This video was created with Opus 👇

Introducing Claude Opus 5.5 for launch videos. Give it an idea, your assets, and a visual reference. Through the Motion MCP, it can build animated scenes, typography, and transitions into a video you can keep editing. Want something changed? Tell Claude, and it updates the video in Motion. This video was created with Opus 👇

Motion

95,816 次观看 • 12 天前

Grok Imagine API just released A world-class video generation + video editing model Text-to-Video: Turn simple prompts into rich video clips with audio Image Generation + Editing: Bring ideas to life with visuals from scratch Video Editing Tools: Restyle scenes, add/remove props, control motion Best-in-Class Quality + Low Latency: Designed to deliver fast, cost-efficient results API pricing: Image input: $0.002 Video input : $0.01 Video output : $0.05

Grok Imagine API just released A world-class video generation + video editing model Text-to-Video: Turn simple prompts into rich video clips with audio Image Generation + Editing: Bring ideas to life with visuals from scratch Video Editing Tools: Restyle scenes, add/remove props, control motion Best-in-Class Quality + Low Latency: Designed to deliver fast, cost-efficient results API pricing: Image input: $0.002 Video input : $0.01 Video output : $0.05

X Freeze

15,078 次观看 • 8 个月前

Track Anything: Segment Anything Meets Videos Track-Anything is a flexible and interactive tool for video object tracking and segmentation suitable for: - Video object tracking and segmentation with shot changes. - Visualized development and data annnotation for video object tracking and segmentation. - Object-centric downstream video tasks, such as video inpainting and editing. abs: github:

Track Anything: Segment Anything Meets Videos Track-Anything is a flexible and interactive tool for video object tracking and segmentation suitable for: - Video object tracking and segmentation with shot changes. - Visualized development and data annnotation for video object tracking and segmentation. - Object-centric downstream video tasks, such as video inpainting and editing. abs: github:

AK

578,691 次观看 • 3 年前

We just shipped video sharing for Eleven v3. You can now turn any generation into a short video. Here's an example.

We just shipped video sharing for Eleven v3. You can now turn any generation into a short video. Here's an example.

ElevenLabs

39,849 次观看 • 1 年前

More examples of what Meta Movie Gen can do across video generation, precise video editing, personalized video generation and audio generation.

More examples of what Meta Movie Gen can do across video generation, precise video editing, personalized video generation and audio generation.

AI at Meta

97,106 次观看 • 2 年前

AI just fired the video editor 3D Motion just dropped the fastest way to turn any image into a cinematic video. With depth mapping, smooth camera moves, and movie-like lighting, it makes static photos look like Hollywood shots. Tutorial below + Prompts:

AI just fired the video editor 3D Motion just dropped the fastest way to turn any image into a cinematic video. With depth mapping, smooth camera moves, and movie-like lighting, it makes static photos look like Hollywood shots. Tutorial below + Prompts:

Robin Delta

137,796 次观看 • 1 年前

We open-source Action Images — a new representation that translates 7-DoF robot actions into interpretable images. Video models are emerging as powerful robotic foundation models, but a key challenge remains: how can we seamlessly integrate robot policies into video models? Instead of representing actions as low-dimensional control tokens, Action Images provide a pixel-grounded action representation, reframing policy learning as a visual tracking problem! By unifying observations and actions in the same video space, Action Images enable a unified robotics world model that supports video-action joint generation, action-conditioned video generation, and action labeling! Code: Paper:

We open-source Action Images — a new representation that translates 7-DoF robot actions into interpretable images. Video models are emerging as powerful robotic foundation models, but a key challenge remains: how can we seamlessly integrate robot policies into video models? Instead of representing actions as low-dimensional control tokens, Action Images provide a pixel-grounded action representation, reframing policy learning as a visual tracking problem! By unifying observations and actions in the same video space, Action Images enable a unified robotics world model that supports video-action joint generation, action-conditioned video generation, and action labeling! Code: Paper:

Chuang Gan

20,809 次观看 • 3 个月前

Creating an e-commerce explainer video usually means juggling a presenter, script, visuals, motion graphics, infographics, voiceover, and editing. With Fotor Video Agent, you can turn one idea into the whole video. AI Presenter for the story. Motion Graphics for key messages. Infographics to make complex e-commerce data easy to understand. Voiceover and subtitles to keep everything in sync. And it doesn’t stop at generating a video. The Motion Graphics are fully editable, so you can fine-tune the text, layout, timing, and visual elements after generation. One Agent. One workflow. From idea to production-ready explainer video. No camera. No After Effects. No complicated editing workflow. Just describe what you want, and let Fotor Video Agent build it.

Creating an e-commerce explainer video usually means juggling a presenter, script, visuals, motion graphics, infographics, voiceover, and editing. With Fotor Video Agent, you can turn one idea into the whole video. AI Presenter for the story. Motion Graphics for key messages. Infographics to make complex e-commerce data easy to understand. Voiceover and subtitles to keep everything in sync. And it doesn’t stop at generating a video. The Motion Graphics are fully editable, so you can fine-tune the text, layout, timing, and visual elements after generation. One Agent. One workflow. From idea to production-ready explainer video. No camera. No After Effects. No complicated editing workflow. Just describe what you want, and let Fotor Video Agent build it.

Fotor

46,584 次观看 • 1 个月前

📢 Lyra: Generative 3D Scene Reconstruction via Video Diffusion Model Self-Distillation Got only one or a few images and wondering if recovering the 3D environment is a reconstruction or generation problem? Why not do it with a generative reconstruction model! We show that a camera-conditioned video diffusion model can be transformed into a generative reconstruction model that directly outputs a high-quality 3D Gaussian Splatting representation through self-distillation, without requiring real-world training data. Check out our results in the video (wait for dynamic scenes in the second half!) : Project Page: Code and Models: Paper:

📢 Lyra: Generative 3D Scene Reconstruction via Video Diffusion Model Self-Distillation Got only one or a few images and wondering if recovering the 3D environment is a reconstruction or generation problem? Why not do it with a generative reconstruction model! We show that a camera-conditioned video diffusion model can be transformed into a generative reconstruction model that directly outputs a high-quality 3D Gaussian Splatting representation through self-distillation, without requiring real-world training data. Check out our results in the video (wait for dynamic scenes in the second half!) : Project Page: Code and Models: Paper:

Sherwin Bahmani

66,897 次观看 • 1 年前

Generated the video using Grok and turned it into 3D Gaussian Splatting.

Generated the video using Grok and turned it into 3D Gaussian Splatting.

Jaskirat Singh

49,096 次观看 • 9 个月前

This video is wrapped up in puns. There's a lot of audio issues, this was mostly to test a node that allows the motion and context information to carry over into the next generation. There are 16 different clips that make up this one video. Can you spot the seams? #minimaxh3

This video is wrapped up in puns. There's a lot of audio issues, this was mostly to test a node that allows the motion and context information to carry over into the next generation. There are 16 different clips that make up this one video. Can you spot the seams? #minimaxh3

_GAi

163,233 次观看 • 1 个月前

byebye expensive motion tracking equipment 👋 ai makes motion capturing so easy now! nvidia presented GENMO last week, a new model that can generate and estimate human motion from text, audio, video, and 3D keyframes

byebye expensive motion tracking equipment 👋 ai makes motion capturing so easy now! nvidia presented GENMO last week, a new model that can generate and estimate human motion from text, audio, video, and 3D keyframes

Dreaming Tulpa 🥓👑

73,693 次观看 • 1 年前

Video editing is hard? Says who. each::labs changed that for me. Generation, editing, everything in one platform. And no, this isn't a dream. Every edit in this video took minutes.

Video editing is hard? Says who. each::labs changed that for me. Generation, editing, everything in one platform. And no, this isn't a dream. Every edit in this video took minutes.

Sharon Riley

82,734 次观看 • 1 个月前

CineMaster A 3D-Aware and Controllable Framework for Cinematic Text-to-Video Generation

CineMaster A 3D-Aware and Controllable Framework for Cinematic Text-to-Video Generation

AK

18,843 次观看 • 1 年前