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。

Live monocular 3D reconstruction from camera feed, with multi-frame stitching and relative positioning (frame buffer of 4)... running on M3 mac cpu at 24fps... visualised using Rerun next up... realtime semantic navigation in 3d space...

ud

1,524 subscribers

22,080 次观看 • 10 个月前 •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

🚀Introducing LLaVA-NeXT Interleave: Now AI can understand and reason with multiple images at once - This opens up multi-image scenarios like multi-frame videos, multi-view 3D, and multiple inter-leaved images. - An all round LMM that can understand videos, images, and 3D More⬇️

🚀Introducing LLaVA-NeXT Interleave: Now AI can understand and reason with multiple images at once - This opens up multi-image scenarios like multi-frame videos, multi-view 3D, and multiple inter-leaved images. - An all round LMM that can understand videos, images, and 3D More⬇️

Gradio

27,665 次观看 • 2 年前

You don’t always need a 3D library to create 3D effects. This paper wave is pure SVG. Every point is projected into 3D space and redrawn frame by frame.

You don’t always need a 3D library to create 3D effects. This paper wave is pure SVG. Every point is projected into 3D space and redrawn frame by frame.

Gabriel

40,957 次观看 • 17 天前

This new world model is insane running realtime at +24fps, this is neither video or 3D, it’s something in-between 🤯 holodeck incoming

This new world model is insane running realtime at +24fps, this is neither video or 3D, it’s something in-between 🤯 holodeck incoming

Linus ✦ Ekenstam

86,576 次观看 • 1 年前

Night City with #RTXOn 😍 Featuring DLSS 4 with Multi Frame Generation and enhanced Ray Reconstruction! 🎥: Rimaeternax

Night City with #RTXOn 😍 Featuring DLSS 4 with Multi Frame Generation and enhanced Ray Reconstruction! 🎥: Rimaeternax

NVIDIA GeForce

85,182 次观看 • 1 年前

Check-out 📢DUSt3R📢 - a new 3D reconstruction model that works with no prior info on camera calibration nor viewpoint poses! Outperforms SoA monocular & multiview depth estimation & relative pose estimation. Paper, demo, videos (& soon code!) available

Check-out 📢DUSt3R📢 - a new 3D reconstruction model that works with no prior info on camera calibration nor viewpoint poses! Outperforms SoA monocular & multiview depth estimation & relative pose estimation. Paper, demo, videos (& soon code!) available

NAVER LABS Europe

17,160 次观看 • 2 年前

Semantically annotating 3D gaussian splats on the fly using gemini 3.1 + sparkjs 1. Load any 3D scene and hit scan 2. Get 2D detections from VLM 3. Cluster outputs & project into 3D world space 4. Save as a persistent 3D semantic layer Inspired by Alexander Chen's experiments with gemini visual intelligence. Just had to try to lift it from 2D to 3D!

Semantically annotating 3D gaussian splats on the fly using gemini 3.1 + sparkjs 1. Load any 3D scene and hit scan 2. Get 2D detections from VLM 3. Cluster outputs & project into 3D world space 4. Save as a persistent 3D semantic layer Inspired by Alexander Chen's experiments with gemini visual intelligence. Just had to try to lift it from 2D to 3D!

Bilawal Sidhu

56,910 次观看 • 5 个月前

Over the holiday period, we took an early look at the new DLSS 4 feature set running on an RTX 5080 in Cyberpunk 2077. Super resolution, multi frame-gen, ray reconstruction - initial thoughts and capture right here:

Over the holiday period, we took an early look at the new DLSS 4 feature set running on an RTX 5080 in Cyberpunk 2077. Super resolution, multi frame-gen, ray reconstruction - initial thoughts and capture right here:

Digital Foundry

254,406 次观看 • 1 年前

Indiana Jones and the Great Circle’s new update features DLSS 4 with Multi Frame Generation, Ray Reconstruction, and fully ray-traced shadows from all light sources. #RTXOn →

Indiana Jones and the Great Circle’s new update features DLSS 4 with Multi Frame Generation, Ray Reconstruction, and fully ray-traced shadows from all light sources. #RTXOn →

NVIDIA GeForce

79,909 次观看 • 1 年前

📢 3D world models from video diffusion suffer from inconsistent frames -> blurry output. Our fix: instead of naïve 3D reconstruction, we non-rigidly align each frame into a globally-consistent 3DGS representation. ->sharp visuals on top of any VDM!

📢 3D world models from video diffusion suffer from inconsistent frames -> blurry output. Our fix: instead of naïve 3D reconstruction, we non-rigidly align each frame into a globally-consistent 3DGS representation. ->sharp visuals on top of any VDM!

Matthias Niessner

53,323 次观看 • 7 个月前

In 2017, Gatorade created the “Water Man” using real water, not CGI, by syncing motion capture with 2,048 nozzles and strobe lights to form a 3D human shape frame by frame.

In 2017, Gatorade created the “Water Man” using real water, not CGI, by syncing motion capture with 2,048 nozzles and strobe lights to form a 3D human shape frame by frame.

Science girl

81,593 次观看 • 9 个月前

Spatio-Temporal Reconstruction Model for Large-Scale Outdoor Scenes Contributions: • We propose STORM, the first feed-forward, self-supervised method for fast and accurate reconstruction of dynamic 3D scenes from sparse, multi-timestep, posed camera images. • Our bottom-up framework aggregates and transforms per-frame 3D Gaussian Splats into a cohesive scene representation, enabling self-supervised motion estimation. Furthermore, we introduce motion tokens that capture common motion primitives and regularize motion predictions, facilitating dynamic motion group segmentation without explicit motion or correspondence supervision. • We present several enhancements for in-the-wild scenarios, including sky modeling, camera exposure inconsistency handling, large novel-view extrapolation, and fine-grained human motions reconstruction, making STORM well-suited for real-world applications.

Spatio-Temporal Reconstruction Model for Large-Scale Outdoor Scenes Contributions: • We propose STORM, the first feed-forward, self-supervised method for fast and accurate reconstruction of dynamic 3D scenes from sparse, multi-timestep, posed camera images. • Our bottom-up framework aggregates and transforms per-frame 3D Gaussian Splats into a cohesive scene representation, enabling self-supervised motion estimation. Furthermore, we introduce motion tokens that capture common motion primitives and regularize motion predictions, facilitating dynamic motion group segmentation without explicit motion or correspondence supervision. • We present several enhancements for in-the-wild scenarios, including sky modeling, camera exposure inconsistency handling, large novel-view extrapolation, and fine-grained human motions reconstruction, making STORM well-suited for real-world applications.

MrNeRF

53,292 次观看 • 1 年前

PixWorld Unifying 3D Scene Generation and Reconstruction in Pixel Space

PixWorld Unifying 3D Scene Generation and Reconstruction in Pixel Space

AK

16,750 次观看 • 4 个月前

Made with Grok Imagine using the last frame of each sequence and using it as the starting frame of the next sequence.

Made with Grok Imagine using the last frame of each sequence and using it as the starting frame of the next sequence.

Déborah

7,954,520 次观看 • 1 年前

💥 Think more real data is needed for scene reconstruction? Think again! Meet MegaSynth: scaling up feed-forward 3D scene reconstruction with synthesized scenes. In 3 days, it generates 700K scenes for training—70x larger than real data! ✨ The secret? Reconstruction is mostly non-semantic! No need to rely heavily on real or highly realistic synthetic data. 🌐 Project: (1/4)

💥 Think more real data is needed for scene reconstruction? Think again! Meet MegaSynth: scaling up feed-forward 3D scene reconstruction with synthesized scenes. In 3 days, it generates 700K scenes for training—70x larger than real data! ✨ The secret? Reconstruction is mostly non-semantic! No need to rely heavily on real or highly realistic synthetic data. 🌐 Project: (1/4)

Hanwen Jiang

27,343 次观看 • 1 年前

A mix of 2D, 3D, stop motion and frame by frame embroidery in this Fortnite x Chappell Roan promotional animation. Directed by Raman Djafari at Blink studio.

A mix of 2D, 3D, stop motion and frame by frame embroidery in this Fortnite x Chappell Roan promotional animation. Directed by Raman Djafari at Blink studio.

Catsuka

43,694 次观看 • 8 个月前

** Sega Genesis 3D engine update 4 ** Never can leave a good rabbit hole undug ! .. updates since last update : Changed frame buffer from a VDP tile friendly buffer to a linear buffer which needs CPU translation , which speedup drawing a LOT not having to worry about tile boundaries albiet slows buffer sending to VRAM. The interior filling of triangles now is running at 1.5c ( 68k cycles ) per pixel ! however sending the buffer is twice as slow per byte as we need the CPU to send the buffer to VRAM to write to the tile format the VDP needs, rather than DMA directly to VRAM . Overall speedup 25% + as the drawing speedup is much greater than the frame buffer slow down. Optimised Line drawing and Triangle filling -near optimal now for the linear frame buffer. Triangle drawing might get a little bit more yet. Quad drawing is next on the list as that will speedup drawing quads. Some gains will come from been smart about partial clearing of RAM buffers and partial sending to VRAM buffers, rather than send the whole buffer. Vertice rotation and projection needs a lot of optimising .. i expect good gains to come from this - this is all in C at present and all calced in real time. Line and Triangle drawing are in assembly but all vertex manipulation is in C. #SGDK #MegaDrive #SegaGenesis

Sega Genesis 3D engine update 4 Never can leave a good rabbit hole undug ! .. updates since last update : Changed frame buffer from a VDP tile friendly buffer to a linear buffer which needs CPU translation , which speedup drawing a LOT not having to worry about tile boundaries albiet slows buffer sending to VRAM. The interior filling of triangles now is running at 1.5c ( 68k cycles ) per pixel ! however sending the buffer is twice as slow per byte as we need the CPU to send the buffer to VRAM to write to the tile format the VDP needs, rather than DMA directly to VRAM . Overall speedup 25% + as the drawing speedup is much greater than the frame buffer slow down. Optimised Line drawing and Triangle filling -near optimal now for the linear frame buffer. Triangle drawing might get a little bit more yet. Quad drawing is next on the list as that will speedup drawing quads. Some gains will come from been smart about partial clearing of RAM buffers and partial sending to VRAM buffers, rather than send the whole buffer. Vertice rotation and projection needs a lot of optimising .. i expect good gains to come from this - this is all in C at present and all calced in real time. Line and Triangle drawing are in assembly but all vertex manipulation is in C. #SGDK #MegaDrive #SegaGenesis

Shannon Birt

15,507 次观看 • 6 个月前

Glob3R: Global Structure-from-Motion With 3D Foundation Models HKUST Spatial Artificial Intelligence Lab · Alibaba Group · Nanjin University · Fudan University Glob3R is a scalable global 3D reconstruction framework built on geometric foundation models. It converts dense feed-forward predictions into reliable multi-view tracks and jointly refines camera poses and scene geometry through motion averaging and bundle adjustment. Glob3R improves reconstruction accuracy and consistency on long sequences, large-scale scenes, and unordered image collections.

Glob3R: Global Structure-from-Motion With 3D Foundation Models HKUST Spatial Artificial Intelligence Lab · Alibaba Group · Nanjin University · Fudan University Glob3R is a scalable global 3D reconstruction framework built on geometric foundation models. It converts dense feed-forward predictions into reliable multi-view tracks and jointly refines camera poses and scene geometry through motion averaging and bundle adjustment. Glob3R improves reconstruction accuracy and consistency on long sequences, large-scale scenes, and unordered image collections.

Ryohei Sasaki@engineer

18,157 次观看 • 3 个月前

💨 Motion accumulation blur. UV is double-buffer animated by direction and scale of camera-relative movement (in screen space).

💨 Motion accumulation blur. UV is double-buffer animated by direction and scale of camera-relative movement (in screen space).

Mirza Beig

20,328 次观看 • 4 个月前

How can lightweight drones without depth cameras navigate using monocular images? Check out our paper at ISER 2023! MonoNav: MAV Navigation via Monocular Depth Estimation and Reconstruction arXiv: website: Work led by Nate Simon

How can lightweight drones without depth cameras navigate using monocular images? Check out our paper at ISER 2023! MonoNav: MAV Navigation via Monocular Depth Estimation and Reconstruction arXiv: website: Work led by Nate Simon

Anirudha Majumdar

27,726 次观看 • 2 年前