Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

Thrilled to introduce Video Depth Anything to support Depth Estimation for super-long videos (over 5 minutes). 👉It enjoys all the benefits of #DepthAnything: high-quality, fast, robust, etc. Proj Page:

bingyikang's profile picture

Bingyi Kang

1,111 subscribers

25,464 Aufrufe • vor 1 Jahr •via X (Twitter)

KunstWissenschaft & TechnologieBildung#DepthAnything
LIVE

Anya Rossi• Live Now

Private livecam show

6 Kommentare

Profilbild von Bingyi Kang
Bingyi Kangvor 1 Jahr

Arxiv: Code: HuggingFace Demo: Thank again for the generous support of @huggingface @pcuenq @_akhaliq in severing our demo.

Profilbild von Suppybird Kingdom
Suppybird Kingdomvor 1 Jahr

Thank you. Very useful <3

Profilbild von Jonathan Granskog
Jonathan Granskogvor 1 Jahr

looks great! congrats!

Profilbild von Aradhye Agarwal
Aradhye Agarwalvor 1 Jahr

Hi, great work! Could you share the link to the RGB videos?

Profilbild von Leor Cohen
Leor Cohenvor 1 Jahr

Looks amazing! 🔥

Profilbild von AIサポートで楽しく小説を書こう。NovelShaft
AIサポートで楽しく小説を書こう。NovelShaftvor 1 Jahr

素晴らしいプロジェクトですね！ビデオの深度推定に興味があります 😊

Ähnliche Videos

Introducing MegaSaM! 🎥 Accurate, fast, & robust structure + camera estimation from casual monocular videos of dynamic scenes! MegaSaM outputs camera parameters and consistent video depth, scaling to long videos with unconstrained camera paths and complex scene dynamics!
2:00

Introducing MegaSaM! 🎥 Accurate, fast, & robust structure + camera estimation from casual monocular videos of dynamic scenes! MegaSaM outputs camera parameters and consistent video depth, scaling to long videos with unconstrained camera paths and complex scene dynamics!

Zhengqi Li

56,992 Aufrufe • vor 1 Jahr

Want to use Depth Anything, but need metric depth rather than relative depth? Thrilled to introduce Prompt Depth Anything, a new paradigm for accurate metric depth estimation with up to 4K resolution. 👉Key Message: Depth foundation models like DA have already internalized rich geometric knowledge of the 3D world but lack a proper way to elicit it. Inspired by the success of prompting in LLMs, we propose prompting Depth Anything with metric cues to produce metric depth. This method proves to be very effective when using a low-cost lidar (e.g., iPhone's LiDAR), which is widely available, as prompts. We believe the prompt can generalize to other forms as long as scale information is provided. Prompt Depth Anything offers 1⃣A series of models for iPhone lidars. 2⃣4D reconstruction from monocular videos (captured with iPhone). 3⃣Improved generalization ability for robot manipulation, e.g. Training on cans but generalizing on glasses. 4⃣More detailed depth annotations for the ScanNet++ dataset. The first author is our excellent intern Haotong Lin. Paper: Huggingface: Project Page: Code:
1:20

Want to use Depth Anything, but need metric depth rather than relative depth? Thrilled to introduce Prompt Depth Anything, a new paradigm for accurate metric depth estimation with up to 4K resolution. 👉Key Message: Depth foundation models like DA have already internalized rich geometric knowledge of the 3D world but lack a proper way to elicit it. Inspired by the success of prompting in LLMs, we propose prompting Depth Anything with metric cues to produce metric depth. This method proves to be very effective when using a low-cost lidar (e.g., iPhone's LiDAR), which is widely available, as prompts. We believe the prompt can generalize to other forms as long as scale information is provided. Prompt Depth Anything offers 1⃣A series of models for iPhone lidars. 2⃣4D reconstruction from monocular videos (captured with iPhone). 3⃣Improved generalization ability for robot manipulation, e.g. Training on cans but generalizing on glasses. 4⃣More detailed depth annotations for the ScanNet++ dataset. The first author is our excellent intern Haotong Lin. Paper: Huggingface: Project Page: Code:

Bingyi Kang

67,604 Aufrufe • vor 1 Jahr

After a year of team work, we're thrilled to introduce Depth Anything 3 (DA3)! 🚀 Aiming for human-like spatial perception, DA3 extends monocular depth estimation to any-view scenarios, including single images, multi-view images, and video. In pursuit of minimal modeling, DA3 reveals two key insights: 💎 A plain transformer (e.g., vanilla DINO) is enough. No specialized architecture. ✨ A single depth-ray representation is enough. No complex 3D tasks. Three series of models have been released: the main DA3 series, a monocular metric estimation series, and a monocular depth estimation series. The core team members, aside from me: Haotong Lin, Sili Chen, Jun Hao Liew, Donny Y. Chen. 👇(1/n) #DepthAnything3
0:32

After a year of team work, we're thrilled to introduce Depth Anything 3 (DA3)! 🚀 Aiming for human-like spatial perception, DA3 extends monocular depth estimation to any-view scenarios, including single images, multi-view images, and video. In pursuit of minimal modeling, DA3 reveals two key insights: 💎 A plain transformer (e.g., vanilla DINO) is enough. No specialized architecture. ✨ A single depth-ray representation is enough. No complex 3D tasks. Three series of models have been released: the main DA3 series, a monocular metric estimation series, and a monocular depth estimation series. The core team members, aside from me: Haotong Lin, Sili Chen, Jun Hao Liew, Donny Y. Chen. 👇(1/n) #DepthAnything3

Bingyi Kang

514,905 Aufrufe • vor 8 Monaten

Extract accurate depth from your 3D gaussians! Fast, accurate, and easy to use. Code of RaDe-GS is released! project page: code: paper: No need to change your 3D GS representations, we propose a play-and-plug rasterizer to extract accurate depth and normal from 3D gaussians, leading to the possibility of high-quality geometry reconstruction. It's quite efficient and fast, achieving 0.69mm errors on DTU dataset in only 5 minutes. Very easy to adopt the rasterizer in your any 3dgs projects.
0:24

Extract accurate depth from your 3D gaussians! Fast, accurate, and easy to use. Code of RaDe-GS is released! project page: code: paper: No need to change your 3D GS representations, we propose a play-and-plug rasterizer to extract accurate depth and normal from 3D gaussians, leading to the possibility of high-quality geometry reconstruction. It's quite efficient and fast, achieving 0.69mm errors on DTU dataset in only 5 minutes. Very easy to adopt the rasterizer in your any 3dgs projects.

Xiao-Xiao Long

31,290 Aufrufe • vor 2 Jahren

Took it further with Omma AI, (Spline ) now it works on video, and everything working in a Web Browser. Drop a video, Depth Anything v2 calculates the depth map frame by frame, and Three.js renders a fully lit 3D mesh in real time. Depth estimation on video in the browser +Automatic baking pass + Dynamic lighting reacting to the geometry #vibecoding #threejs #webgl
0:40

Took it further with Omma AI, (Spline ) now it works on video, and everything working in a Web Browser. Drop a video, Depth Anything v2 calculates the depth map frame by frame, and Three.js renders a fully lit 3D mesh in real time. Depth estimation on video in the browser +Automatic baking pass + Dynamic lighting reacting to the geometry #vibecoding #threejs #webgl

Joseph Azar

147,666 Aufrufe • vor 2 Monaten

Here’s a sneak peek using Rerun and Gradio for data annotation. It uses Video Depth Anything and Segment Anything 2 under the hood to generate segmentation masks and depth maps/point clouds. More to share next week.
0:30

Here’s a sneak peek using Rerun and Gradio for data annotation. It uses Video Depth Anything and Segment Anything 2 under the hood to generate segmentation masks and depth maps/point clouds. More to share next week.

Pablo Vela

36,719 Aufrufe • vor 1 Jahr

ViPE: Video Pose Engine for 3D Geometric Perception Contributions: • A robust and efficient framework, ViPE, for estimating camera parameters and dense depth from diverse, in-the-wild videos. • A system design that integrates the strengths of classical SLAM (efficiency, scalability) and learned models (robustness), with key improvements in efficiency, dynamic object handling, and depth quality over prior work. • A large-scale dataset of annotated videos, created using ViPE, to facilitate future research in 3D computer vision.
0:22

ViPE: Video Pose Engine for 3D Geometric Perception Contributions: • A robust and efficient framework, ViPE, for estimating camera parameters and dense depth from diverse, in-the-wild videos. • A system design that integrates the strengths of classical SLAM (efficiency, scalability) and learned models (robustness), with key improvements in efficiency, dynamic object handling, and depth quality over prior work. • A large-scale dataset of annotated videos, created using ViPE, to facilitate future research in 3D computer vision.

MrNeRF

42,553 Aufrufe • vor 11 Monaten

Introducing Modality Forcing, a recipe for post-training T2I models for SOTA RGB-Depth generation! Text-to-image (T2I) models learn rich representations of the spatial world. How do we build on this prior for high-quality depth generation? 🧵 [1/6]
0:25

Introducing Modality Forcing, a recipe for post-training T2I models for SOTA RGB-Depth generation! Text-to-image (T2I) models learn rich representations of the spatial world. How do we build on this prior for high-quality depth generation? 🧵 [1/6]

Bardienus Duisterhof

64,543 Aufrufe • vor 1 Monat

Happy Chinese New Year, and thank you all for your love and support over the past year. On the last day of the Chinese Lunar Calendar year of 2024, we are thrilled to introduce to you the Blender plugin for Hunyun3D 2.0. Enjoy yourself!
1:17

Happy Chinese New Year, and thank you all for your love and support over the past year. On the last day of the Chinese Lunar Calendar year of 2024, we are thrilled to introduce to you the Blender plugin for Hunyun3D 2.0. Enjoy yourself!

Tencent Hy

82,805 Aufrufe • vor 1 Jahr

Working on room scan streaming for the Lasertag game: headsets build a shared heightmap of the playspace collaboratively in the background. Quest's depth estimation is only good up to ~5 meters, so this lets people fight at greater distances
0:43

Working on room scan streaming for the Lasertag game: headsets build a shared heightmap of the playspace collaboratively in the background. Quest's depth estimation is only good up to ~5 meters, so this lets people fight at greater distances

jt

14,474 Aufrufe • vor 1 Jahr

New Video! Excited to share our work on fast view synthesis of casual videos. 🤩 high-quality dynamic view synthesis 🤩 applicable for a wide range of casual videos 🤩 fast training and rendering
1:02

New Video! Excited to share our work on fast view synthesis of casual videos. 🤩 high-quality dynamic view synthesis 🤩 applicable for a wide range of casual videos 🤩 fast training and rendering

Jia-Bin Huang

27,360 Aufrufe • vor 2 Jahren

Monica Long | Benefits of Ripple $ Stablecoin 1. Reduces volatility of native crypto for transacting 2. Uses a global blockchain rail 24x7 4. Nearly free 👉Expect Ripple to keep innovating for #XRPL. 👉Next challenge: Create breadth and depth of liquidity in key currency pairs.
2:41

Monica Long | Benefits of Ripple $ Stablecoin 1. Reduces volatility of native crypto for transacting 2. Uses a global blockchain rail 24x7 4. Nearly free 👉Expect Ripple to keep innovating for #XRPL. 👉Next challenge: Create breadth and depth of liquidity in key currency pairs.

🌸Eri ~ Carpe Diem

18,593 Aufrufe • vor 11 Monaten

SEI creator fund round 5 voting is live! This round 5 is all about creating visual content (short/long videos, livestream, etc...) What I will be doing in this round 5 - Tutorials - Livestreams - Entertaining videos These are the tools I propose using the fund for: - A good camera - Lights - Quality wireless/table mic - A Laptop for video editing With y'all support, all of these will be achievable... Support me with a vote ✅ How to vote: - Fund your $SEI wallet - Go to - Add my project/project to cart - Go to your cart - Input the amount of $sei you want to donate (minimum of $1) - Confirm your donation Done ✅
1:13

SEI creator fund round 5 voting is live! This round 5 is all about creating visual content (short/long videos, livestream, etc...) What I will be doing in this round 5 - Tutorials - Livestreams - Entertaining videos These are the tools I propose using the fund for: - A good camera - Lights - Quality wireless/table mic - A Laptop for video editing With y'all support, all of these will be achievable... Support me with a vote ✅ How to vote: - Fund your $SEI wallet - Go to - Add my project/project to cart - Go to your cart - Input the amount of $sei you want to donate (minimum of $1) - Confirm your donation Done ✅

MhiztaL⭕EL 💎 🔴

12,459 Aufrufe • vor 1 Jahr

full video, $10 hardware, this was just a proof of concept as well. Its going to get a lot better. If your research lab or startup wants wide angle high quality depth for cheap, message me.
0:23

full video, $10 hardware, this was just a proof of concept as well. Its going to get a lot better. If your research lab or startup wants wide angle high quality depth for cheap, message me.

brian-machado-high-inference

422,958 Aufrufe • vor 9 Monaten

S&P 500 $SPX book depth is still extremely thin compared to July 2025 and 2024. Remember, book depth is the amount of open orders on either side of the current price. Basically if book depth is wide it tends to support the market because moves are absorbed by those outstanding bids. When book depth is thin there's less cushion to absorb large moves. That's when you can see accelerated shifts if there's a trigger. With OpEx now over and VIXpiration on Wednesday it will be interesting to see what happens to liquidity and this chart.
12:41

S&P 500 $SPX book depth is still extremely thin compared to July 2025 and 2024. Remember, book depth is the amount of open orders on either side of the current price. Basically if book depth is wide it tends to support the market because moves are absorbed by those outstanding bids. When book depth is thin there's less cushion to absorb large moves. That's when you can see accelerated shifts if there's a trigger. With OpEx now over and VIXpiration on Wednesday it will be interesting to see what happens to liquidity and this chart.

Financelot

27,158 Aufrufe • vor 2 Tagen

We used to take music like this for granted back in the 70’s and 80’s. Now, when you hear it decades later, it seems to have an extra depth of colour and quality…
3:00

We used to take music like this for granted back in the 70’s and 80’s. Now, when you hear it decades later, it seems to have an extra depth of colour and quality…

🌸🎵 Beautiful Melody 🎶💖

26,940 Aufrufe • vor 2 Monaten

Some video (from Dan Kraker ) of the Lester River in Duluth running high and fast thanks to 5"+ of rain over the weekend.
0:21

Some video (from Dan Kraker ) of the Lester River in Duluth running high and fast thanks to 5"+ of rain over the weekend.

Cathy Wurzer

32,099 Aufrufe • vor 2 Jahren

It feels profoundly weird to not be working on Black Ops 7 In Depth this year. I used to make these kind of videos every day, for over a decade. So it feels like a part of me is missing.
10:38

It feels profoundly weird to not be working on Black Ops 7 In Depth this year. I used to make these kind of videos every day, for over a decade. So it feels like a part of me is missing.

Drift0r

61,356 Aufrufe • vor 8 Monaten

After a year of Design & Development, I'm thrilled to introduce The Pillhead 📦 This limited collection of 111 high-quality resin figurines is my first-ever physical drop, ready to be shipped. Fill out the form below to secure your spot! This is a milestone for me, and I'm honored to have you on this journey. Thank you for your support. #IPWT🥂 With 💜 CFW
1:19

After a year of Design & Development, I'm thrilled to introduce The Pillhead 📦 This limited collection of 111 high-quality resin figurines is my first-ever physical drop, ready to be shipped. Fill out the form below to secure your spot! This is a milestone for me, and I'm honored to have you on this journey. Thank you for your support. #IPWT🥂 With 💜 CFW

CFW •ᴗ•

18,435 Aufrufe • vor 2 Jahren

The team at NVIDIA AI Developer really nailed it with ViPE!! This SLAM system generated pose, depth, point clouds, etc, in a few seconds! Now to get the output to work with GEN3C... If you are interested in running this yourself, it took about 5 minutes to get up and running from scratch.
0:28

The team at NVIDIA AI Developer really nailed it with ViPE!! This SLAM system generated pose, depth, point clouds, etc, in a few seconds! Now to get the output to work with GEN3C... If you are interested in running this yourself, it took about 5 minutes to get up and running from scratch.

Jonathan Stephens

34,947 Aufrufe • vor 11 Monaten