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。

Test3R Learning to Reconstruct 3D at Test Time

AK

501,188 subscribers

55,287 次观看 • 1 年前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

We're rapidly applying lessons learned from Ukraine to enhance U.S. Army readiness. From 3D-printed drones to empowering units to test new tech, we're iterating & learning in real-time to ensure battlefield dominance. #ArmyTransformation #Ukraine #Innovation #testimony

We're rapidly applying lessons learned from Ukraine to enhance U.S. Army readiness. From 3D-printed drones to empowering units to test new tech, we're iterating & learning in real-time to ensure battlefield dominance. #ArmyTransformation #Ukraine #Innovation #testimony

Secretary of the Army

29,461 次观看 • 1 年前

3D Gaussian Splats are cool, but they're static (Part 25). DG-Mesh can reconstruct high-quality and time-consistent 3D meshes from a single video and track mesh vertices over time, which enables texture editing on dynamic objects.

3D Gaussian Splats are cool, but they're static (Part 25). DG-Mesh can reconstruct high-quality and time-consistent 3D meshes from a single video and track mesh vertices over time, which enables texture editing on dynamic objects.

Dreaming Tulpa 🥓👑

79,033 次观看 • 2 年前

men will literally do anything to avoid learning 3d

men will literally do anything to avoid learning 3d

Dreaming Tulpa 🥓👑

51,180 次观看 • 1 年前

🚀 Just dropped the code: #InstantSplat! Reconstruct your 3D scenes in seconds. Get it now at and start building! ✨ #3DVision #AI #ComputerVision #NVlabs #OpenSource

🚀 Just dropped the code: #InstantSplat! Reconstruct your 3D scenes in seconds. Get it now at and start building! ✨ #3DVision #AI #ComputerVision #NVlabs #OpenSource

Zhiwen(Aaron) Fan

89,142 次观看 • 1 年前

Jonas Hübotter is a doctoral Researcher at ETH Zurich. Where he works on Active Fine-Tuning and Local Learning. His paper Transductive Active Learning: Theory and Applications is published at NeurIPS24 and we discuss Efficiently Learning at Test-Time: Active Fine-Tuning of LLMs (submitted for ICLR 25) - live on MLST now.

Jonas Hübotter is a doctoral Researcher at ETH Zurich. Where he works on Active Fine-Tuning and Local Learning. His paper Transductive Active Learning: Theory and Applications is published at NeurIPS24 and we discuss Efficiently Learning at Test-Time: Active Fine-Tuning of LLMs (submitted for ICLR 25) - live on MLST now.

Machine Learning Street Talk

10,246 次观看 • 1 年前

It doesn’t have to be this way! These are the books we at Classic Learning Test use to assess college-bound students.

It doesn’t have to be this way! These are the books we at Classic Learning Test use to assess college-bound students.

Michael Torres

24,615 次观看 • 1 个月前

Unlike machine learning, which separates training from testing, humans learn continuously from day one. “We need to figure out how to ensure AI can learn continuously during the test time”, says Yejin Choi (Stanford University) at #WEF26.

Unlike machine learning, which separates training from testing, humans learn continuously from day one. “We need to figure out how to ensure AI can learn continuously during the test time”, says Yejin Choi (Stanford University) at #WEF26.

World Economic Forum

59,357 次观看 • 4 个月前

🎮 React Three Game on Three.js Resources Learn how to build 3D games using React Three Fiber and #threejs. A great starting point for understanding game logic, interactivity, and real-time 3D on the web. 👉 #threejs #WebGL #GameDev #ReactThreeFiber #3D #Learning

🎮 React Three Game on Three.js Resources Learn how to build 3D games using React Three Fiber and #threejs. A great starting point for understanding game logic, interactivity, and real-time 3D on the web. 👉 #threejs #WebGL #GameDev #ReactThreeFiber #3D #Learning

Three.js Resources

15,838 次观看 • 4 个月前

Check out CAT3D! Image(s)-to-3D in 1 minute! Given any number of real or generated images, CAT3D uses a multi-view diffusion prior to create consistent novel views. These views are used to reconstruct a 3D scene using NeRF/3DGS.

Check out CAT3D! Image(s)-to-3D in 1 minute! Given any number of real or generated images, CAT3D uses a multi-view diffusion prior to create consistent novel views. These views are used to reconstruct a 3D scene using NeRF/3DGS.

Philipp Henzler

12,737 次观看 • 2 年前

POV: You haven't stopped learning 3D

POV: You haven't stopped learning 3D

Jan

76,812 次观看 • 1 年前

How do we reconstruct a 3D scene from photos with varying exposures? Standard methods often fail, leaving you with blown-out colors or disturbing shadows. We're excited to introduce Neural Exposure Fields (NExF), our new work accepted at #NeurIPS2025! 🧵

How do we reconstruct a 3D scene from photos with varying exposures? Standard methods often fail, leaving you with blown-out colors or disturbing shadows. We're excited to introduce Neural Exposure Fields (NExF), our new work accepted at #NeurIPS2025! 🧵

Michael Niemeyer

131,880 次观看 • 8 个月前

Nature presents a captivating confluence of similarity and diversity. Our new method 3D-Fauna learns a pan-category articulated 3D model of quadruped animals from Internet photos. At test time, it turns a single image into an animatable textured 3D mesh in a feed-forward pass.

Nature presents a captivating confluence of similarity and diversity. Our new method 3D-Fauna learns a pan-category articulated 3D model of quadruped animals from Internet photos. At test time, it turns a single image into an animatable textured 3D mesh in a feed-forward pass.

Elliott / Shangzhe Wu

107,220 次观看 • 2 年前

A day full of learning 📈 Felix Rosenqvist details the No. 60 team's test at Iowa Speedway.

A day full of learning 📈 Felix Rosenqvist details the No. 60 team's test at Iowa Speedway.

NTT INDYCAR SERIES

16,020 次观看 • 11 个月前

Napkin sketch to interactive 3D rendering! Antigravity is the agentic development platform for developers to turn hand-drawn sketches into fully functional 3D learning modules.

Napkin sketch to interactive 3D rendering! Antigravity is the agentic development platform for developers to turn hand-drawn sketches into fully functional 3D learning modules.

Google Antigravity

50,303 次观看 • 4 天前

Time to start learning Japanese.

Time to start learning Japanese.

TacticalStNick

372,195 次观看 • 2 个月前

I'm learning a lot from this test

I'm learning a lot from this test

The Matt Walsh Show

74,949 次观看 • 5 个月前

For 3D pose some use different keypoints, others SMPL and other models. It's a mess! With Neural Localizer Fields, we can choose the output at test time! allowing to train using any. Results are real time and SOTA across the board. István Sárándi

For 3D pose some use different keypoints, others SMPL and other models. It's a mess! With Neural Localizer Fields, we can choose the output at test time! allowing to train using any. Results are real time and SOTA across the board. István Sárándi

Gerard Pons-Moll

43,472 次观看 • 1 年前

Sneako opens up to his viewers about learning to protect his time at all costs and why time is the most valuable currency 💰⌛️

Sneako opens up to his viewers about learning to protect his time at all costs and why time is the most valuable currency 💰⌛️

Sneako Updates

29,876 次观看 • 11 个月前

Morw test to more fine-tune the ClayPencil tool, to convert Grease Pencil animations into 3D characters in real time. This time using a reference from the lovely @pedroalpera #Blender #B3D #ClayPencil #GreasePencil #stopmotion #2D #2danimation

Morw test to more fine-tune the ClayPencil tool, to convert Grease Pencil animations into 3D characters in real time. This time using a reference from the lovely @pedroalpera #Blender #B3D #ClayPencil #GreasePencil #stopmotion #2D #2danimation

Daniel Martínez Lara 🔶

82,127 次观看 • 1 年前