Loading video...

Video Failed to Load

Go Home

Some more impressive 3D reconstruction examples from #DUSt3R kindly prepared by our boss ;-) Martin Humenberger NAVER LABS Europe

JeromeRevaud's profile picture

Jerome Revaud

1,673 subscribers

11,043 views • 2 years ago •via X (Twitter)

#DUSt3R
LIVE

Anya Rossi• Live Now

Private livecam show

3 Comments

Jerome Revaud's profile picture
Jerome Revaud2 years ago

(the castle pic (second image pair) shows our actual Lab! )

BenderDrummer's profile picture
BenderDrummer2 years ago

@martinhu @naverlabseurope Even recent and historic lunar lander images would be a really interesting thing to recreate.. with @ILOA_Hawaii for example and the Odysseus mission

Jerome Revaud's profile picture
Jerome Revaud2 years ago

@martinhu @naverlabseurope @ILOA_Hawaii Yes, that would be interesting to see if earthly priors make sense on the moon too. 🌚 I guess it would, to some extent, because terrain on the moon looks like desert on earth…

Related Videos

Instant Video-to-3D with DUST3R Dust3r generates a whole 3D scene from just a couple of images. What if it could: 1. Accept a VIDEO 2. Extract the video frames 3. Turn them into 3D? So added a Gradio Video Component. Here's me generating 3D from a video cc: NAVER LABS Europe
1:13

Instant Video-to-3D with DUST3R Dust3r generates a whole 3D scene from just a couple of images. What if it could: 1. Accept a VIDEO 2. Extract the video frames 3. Turn them into 3D? So added a Gradio Video Component. Here's me generating 3D from a video cc: NAVER LABS Europe

cocktail peanut

41,038 views • 2 years ago

A new model for human mesh recovery, high-performing and w/o using any 3D scans, has been published by my excellent colleagues at NAVER LABS Europe . Excellent work!
0:20

A new model for human mesh recovery, high-performing and w/o using any 3D scans, has been published by my excellent colleagues at NAVER LABS Europe . Excellent work!

Christian Wolf (🦋🦋🦋)

43,782 views • 9 months ago

Marble is the first product from World Labs and is powered by our multimodal world model, which lets anyone create high-fidelity, persistent 3D worlds from just a single image, video, text prompt, or 3D layout. Read more at
0:51

Marble is the first product from World Labs and is powered by our multimodal world model, which lets anyone create high-fidelity, persistent 3D worlds from just a single image, video, text prompt, or 3D layout. Read more at

World Labs

117,310 views • 9 months ago

📢Pixel3DMM: Versatile Screen-Space Priors for Single-Image 3D Face Reconstruction📢 -> highly accurate face reconstruction by training powerful VITs via surface normals and UV-coordinates estimation. The geometric cues from our 2D foundation model backbone constrain the 3DMM parameters, which allows us to achieve remarkable reconstruction accuracy - works for both single image and videos! In addition, we introduce a new 3D face reconstruction benchmark that evaluates both neutral and posed face geometry. 🌍 📷 Great work by Simon Giebenhain Tobias Kirschstein Martin Rünz Lourdes Agapito
0:32

📢Pixel3DMM: Versatile Screen-Space Priors for Single-Image 3D Face Reconstruction📢 -> highly accurate face reconstruction by training powerful VITs via surface normals and UV-coordinates estimation. The geometric cues from our 2D foundation model backbone constrain the 3DMM parameters, which allows us to achieve remarkable reconstruction accuracy - works for both single image and videos! In addition, we introduce a new 3D face reconstruction benchmark that evaluates both neutral and posed face geometry. 🌍 📷 Great work by Simon Giebenhain Tobias Kirschstein Martin Rünz Lourdes Agapito

Matthias Niessner

62,269 views • 1 year ago

🚨 Paper Alert Our recent breakthrough CAST: Component-Aligned 3D Scene Reconstruction from an RGB Image has been accepted by ACM SIGGRAPH 2025 Journal Track! CAST will change the way create scenes in 3D Art and Embody AI. 🚀Soon available at 👇Details
3:41

🚨 Paper Alert Our recent breakthrough CAST: Component-Aligned 3D Scene Reconstruction from an RGB Image has been accepted by ACM SIGGRAPH 2025 Journal Track! CAST will change the way create scenes in 3D Art and Embody AI. 🚀Soon available at 👇Details

Deemos

34,110 views • 1 year ago

Combine Grok Imagine with Topaz Labs Astra. This keeps your creation intact but upgrades it to 60FPS and 1080p as a minimum to bring out more detail and motion. Astra works fully in your browser too. Some quick side-by-side examples.
0:24

Combine Grok Imagine with Topaz Labs Astra. This keeps your creation intact but upgrades it to 60FPS and 1080p as a minimum to bring out more detail and motion. Astra works fully in your browser too. Some quick side-by-side examples.

Jerrod Lew

35,809 views • 10 months ago

By popular demand, we're sharing more examples that showcase CSM Cube's ability to generate high-quality 3D mesh topology and distinct parts. The results shown were achieved using our single-click Image-to-3D feature, but users can obtain even better quality with careful prompting and editing.
0:46

By popular demand, we're sharing more examples that showcase CSM Cube's ability to generate high-quality 3D mesh topology and distinct parts. The results shown were achieved using our single-click Image-to-3D feature, but users can obtain even better quality with careful prompting and editing.

Common Sense Machines

16,904 views • 1 year ago

The biggest Q people want me to ask is 'why the hell do we pay more for energy than anyone else in Europe?' A snip from full Martin v Ofgem boss new episode of the Martin Lewis Podcast. Listen to it all on BBC Sounds, Apple, Spotify & all the usual pod players.
1:20

The biggest Q people want me to ask is 'why the hell do we pay more for energy than anyone else in Europe?' A snip from full Martin v Ofgem boss new episode of the Martin Lewis Podcast. Listen to it all on BBC Sounds, Apple, Spotify & all the usual pod players.

Martin Lewis

527,264 views • 9 months ago

Upgrade your expressive 3D human avatars from #SMPL-X to #SUPR, our latest and greatest body model. SUPR is trained from 1.2M 3D scans, is more expressive, and includes feet with articulation and compression. Code by @NeelayShah8, video by Anastasios Yiannakidis.
0:40

Upgrade your expressive 3D human avatars from #SMPL-X to #SUPR, our latest and greatest body model. SUPR is trained from 1.2M 3D scans, is more expressive, and includes feet with articulation and compression. Code by @NeelayShah8, video by Anastasios Yiannakidis.

Michael Black

60,568 views • 3 years ago

📢📢 𝐀𝐯𝐚𝐭𝟑𝐫 📢📢 Avat3r creates high-quality 3D head avatars from just a few input images in a single forward pass with a new dynamic 3DGS reconstruction model. Video: Project: Our core idea is to make Gaussian Reconstruction Models animatable. We find that a simple cross-attention to an expression code sequence is already sufficient to model complex facial expressions. We then incorporate position maps from DUSt3R and feature maps from Sapiens to facilitate the prediction task. While DUSt3R's position maps act as a pixel-aligned initialization for the Gaussians' positions, the Sapiens feature maps help the cross-view transformer to match corresponding image tokens in the 4 input images. One major challenge in creating a 3D head avatar from smartphone images comes from inconsistent facial expressions when the subject could not remain perfectly static during the capture. We eliminate this static requirement by simply showing our model input images with different facial expressions during training. This technique makes our model robust to inconsistent input images later on. Finally, we show that despite the model has been trained with 4 input images, one can even create a 3D head avatar when only a single image is available. To achieve this, we employ a pre-trained 3D GAN to lift the single image to 3D and then render the 4 input images for our model. This allows us to create 3D head avatars from single images and even highly out-of-distribution examples like AI generated faces, paintings or statues. Great work by Tobias Kirschstein from his internship at Meta with Javier Romero, Artem Sevastopolsky, and Shunsuke Saito
0:38

📢📢 𝐀𝐯𝐚𝐭𝟑𝐫 📢📢 Avat3r creates high-quality 3D head avatars from just a few input images in a single forward pass with a new dynamic 3DGS reconstruction model. Video: Project: Our core idea is to make Gaussian Reconstruction Models animatable. We find that a simple cross-attention to an expression code sequence is already sufficient to model complex facial expressions. We then incorporate position maps from DUSt3R and feature maps from Sapiens to facilitate the prediction task. While DUSt3R's position maps act as a pixel-aligned initialization for the Gaussians' positions, the Sapiens feature maps help the cross-view transformer to match corresponding image tokens in the 4 input images. One major challenge in creating a 3D head avatar from smartphone images comes from inconsistent facial expressions when the subject could not remain perfectly static during the capture. We eliminate this static requirement by simply showing our model input images with different facial expressions during training. This technique makes our model robust to inconsistent input images later on. Finally, we show that despite the model has been trained with 4 input images, one can even create a 3D head avatar when only a single image is available. To achieve this, we employ a pre-trained 3D GAN to lift the single image to 3D and then render the 4 input images for our model. This allows us to create 3D head avatars from single images and even highly out-of-distribution examples like AI generated faces, paintings or statues. Great work by Tobias Kirschstein from his internship at Meta with Javier Romero, Artem Sevastopolsky, and Shunsuke Saito

Matthias Niessner

74,763 views • 1 year ago

We've seen some impressive sends from the 2025 rookies! 👏💨 Head to our YouTube channel to see the best rookie moments of the season from onboard, presented by Qatar Airways ⏯️ #F1 #DrivenToPerfection
0:36

We've seen some impressive sends from the 2025 rookies! 👏💨 Head to our YouTube channel to see the best rookie moments of the season from onboard, presented by Qatar Airways ⏯️ #F1 #DrivenToPerfection

Formula 1

184,075 views • 7 months ago

Spatial computing that's built for the real world, not built to pull you away from it. Here are some examples of what's possible on SPECS with the Lenses built by our developer community.
1:27

Spatial computing that's built for the real world, not built to pull you away from it. Here are some examples of what's possible on SPECS with the Lenses built by our developer community.

SPECS

14,030 views • 1 month ago

A thing of beauty 😲 Enjoy some more of our classic goals at Newcastle from years gone by 👇
0:23

A thing of beauty 😲 Enjoy some more of our classic goals at Newcastle from years gone by 👇

Arsenal

120,160 views • 10 months ago

Don't slander our local dogs, please. Even though some dogs are better securities than others, you cannot judge an entire breed by the behaviour of a single dog. Here are some examples of similar behaviour you complained about, from foreign dogs.
1:12

Don't slander our local dogs, please. Even though some dogs are better securities than others, you cannot judge an entire breed by the behaviour of a single dog. Here are some examples of similar behaviour you complained about, from foreign dogs.

Arojinle

91,754 views • 5 months ago

📢📢 𝐏𝐞𝐫𝐜𝐇𝐞𝐚𝐝: 𝐏𝐞𝐫𝐜𝐞𝐩𝐭𝐮𝐚𝐥 𝐇𝐞𝐚𝐝 𝐌𝐨𝐝𝐞𝐥 𝐟𝐨𝐫 𝐒𝐢𝐧𝐠𝐥𝐞-𝐈𝐦𝐚𝐠𝐞 𝟑𝐃 𝐇𝐞𝐚𝐝 𝐑𝐞𝐜𝐨𝐧𝐬𝐭𝐫𝐮𝐜𝐭𝐢𝐨𝐧 & 𝐄𝐝𝐢𝐭𝐢𝐧𝐠📢📢 PercHead reconstructs realistic 3D heads from a single image and enables disentangled 3D editing via geometric controls and style inputs from images or text. At its core is a generalized 3D head decoder trained with perceptual supervision from DINOv2 and SAM 2.1. We find that our new perceptual loss formulation improves reconstruction fidelity compared to commonly-used methods such as LPIPS. Our trained reconstruction model is able to generate 3D-consistent heads from a single input image. Even with challenging side-view inputs, the model robustly infers missing regions for a coherent, high-fidelity output. In addition, our architecture seamlessly adapts to downstream tasks: by swapping the encoder, we can transform the model into a disentangled 3D editing pipeline. In this scenario, we can control geometry through - potentially hand-drawn - segmentation maps, and condition style via image or text prompt. We also provide an interactive GUI to enable the exploration of our editing pipeline. 🌍 📽️ Great work by Antonio Oroz and Tobias Kirschstein
0:30

📢📢 𝐏𝐞𝐫𝐜𝐇𝐞𝐚𝐝: 𝐏𝐞𝐫𝐜𝐞𝐩𝐭𝐮𝐚𝐥 𝐇𝐞𝐚𝐝 𝐌𝐨𝐝𝐞𝐥 𝐟𝐨𝐫 𝐒𝐢𝐧𝐠𝐥𝐞-𝐈𝐦𝐚𝐠𝐞 𝟑𝐃 𝐇𝐞𝐚𝐝 𝐑𝐞𝐜𝐨𝐧𝐬𝐭𝐫𝐮𝐜𝐭𝐢𝐨𝐧 & 𝐄𝐝𝐢𝐭𝐢𝐧𝐠📢📢 PercHead reconstructs realistic 3D heads from a single image and enables disentangled 3D editing via geometric controls and style inputs from images or text. At its core is a generalized 3D head decoder trained with perceptual supervision from DINOv2 and SAM 2.1. We find that our new perceptual loss formulation improves reconstruction fidelity compared to commonly-used methods such as LPIPS. Our trained reconstruction model is able to generate 3D-consistent heads from a single input image. Even with challenging side-view inputs, the model robustly infers missing regions for a coherent, high-fidelity output. In addition, our architecture seamlessly adapts to downstream tasks: by swapping the encoder, we can transform the model into a disentangled 3D editing pipeline. In this scenario, we can control geometry through - potentially hand-drawn - segmentation maps, and condition style via image or text prompt. We also provide an interactive GUI to enable the exploration of our editing pipeline. 🌍 📽️ Great work by Antonio Oroz and Tobias Kirschstein

Matthias Niessner

18,855 views • 9 months ago

Bleach fans worldwide have stood by the series for decades. We kindly urge the organizers to extend the pre screening events to more international locations beyond Japan and USA. Let’s make some noise and ensure our voice is heard #BleachGlobalTheatre
0:30

Bleach fans worldwide have stood by the series for decades. We kindly urge the organizers to extend the pre screening events to more international locations beyond Japan and USA. Let’s make some noise and ensure our voice is heard #BleachGlobalTheatre

Zeus

16,680 views • 3 months ago

Sharing some more glimpses from my Karnataka pravas. Deeply moved by the warmth of the people and the unmatched energy of our karyakartas.
0:58

Sharing some more glimpses from my Karnataka pravas. Deeply moved by the warmth of the people and the unmatched energy of our karyakartas.

Nitin Nabin

10,768 views • 2 months ago

Most digital twin investments stall for one reason: They’re not grounded in how the world actually looks today. Niantic Spatial’s Reconstruction capability fixes this, creating a living, machine-readable 3D model that stays in sync with reality, so every system, team, and workflow operates from the same ground truth. In a new blog by Trista Pierce, Business Development Lead, explore how Scaniverse’s Reconstruction technology changes the cadence entirely. Read more:
0:35

Most digital twin investments stall for one reason: They’re not grounded in how the world actually looks today. Niantic Spatial’s Reconstruction capability fixes this, creating a living, machine-readable 3D model that stays in sync with reality, so every system, team, and workflow operates from the same ground truth. In a new blog by Trista Pierce, Business Development Lead, explore how Scaniverse’s Reconstruction technology changes the cadence entirely. Read more:

Niantic Spatial 🌎

418,271 views • 4 months ago

Russian woman returned back from Europe and was shocked by Putin's "stability" — grocery prices are soaring into orbit, almost on a daily basis. "Judging by the prices in our 'Pyaterochkas,' it's like I never even came back from Europe: prices are practically the same, and some things abroad are way cheaper than here." And this isn't some expensive store, but the most ordinary store, where everyone goes for the most basic groceries.
0:53

Russian woman returned back from Europe and was shocked by Putin's "stability" — grocery prices are soaring into orbit, almost on a daily basis. "Judging by the prices in our 'Pyaterochkas,' it's like I never even came back from Europe: prices are practically the same, and some things abroad are way cheaper than here." And this isn't some expensive store, but the most ordinary store, where everyone goes for the most basic groceries.

Churchill MGᵀᴹ 🇺🇦🇨🇦🇬🇧🇪🇺🇺🇸

73,545 views • 12 days ago