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。

Looks like Flutter is slowly getting better and better at supporting 3D rendering! This PR adds support for rendering 3D models directly in Flutter: Video and PR are done by Brandon DeRosier, keep up the awesome work 🤩

wolfenrain 💙🏳️‍🌈

7,163 subscribers

86,470 次观看 • 3 年前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

10 条评论

Brandon DeRosier 的头像

Brandon DeRosier3 年前

Thanks for the shout out! This is an early experiment to integrate Impeller Scene in the framework. With the advent of Impeller, it's easier than ever to rapidly build out high value rendering features like this. IMO, Flutter is just getting started. The future is bright!

Stephan E.G. Veenstra 💙 的头像

Stephan E.G. Veenstra 💙3 年前

@algebrandon Great, more things that can distract me from my actual tasks.😅

Rutvik Tak 💙 的头像

Rutvik Tak 💙3 年前

@algebrandon This may just let me get back working with @Blender 😍 Super hyped to check on this🙌

Paras Jain 💙 的头像

Paras Jain 💙3 年前

@_eseidel @algebrandon 🙌

Godwin of the future 的头像

Godwin of the future3 年前

@onuoha_ifeanyi @algebrandon Flutter may just be the next best language after JavaScript. I love it!!

un vato.eth 的头像

un vato.eth3 年前

@algebrandon Eso lo hace three js

Ravin Sharma 的头像

Ravin Sharma3 年前

@algebrandon My gawd. This is so so good. Any docs to try this out. I'm gonna try it in an app.

wolfenra.in 🦄💙🏳️‍🌈 的头像

wolfenra.in 🦄💙🏳️‍🌈3 年前

@algebrandon This got merged to master just a few hours ago, it is nowhere near the point that you can actively use it in your apps

czar 的头像

czar3 年前

@algebrandon I’m currently investing my time into 3Ds using Flutter…

wolfenra.in 🦄💙🏳️‍🌈 的头像

wolfenra.in 🦄💙🏳️‍🌈3 年前

@algebrandon It will be time for it soon 🫡

相关视频

Flutter 💙 + Flame 🔥 = 3D?? What you see in the video is rendered by Flutter, no native packages just Flutter & Flame! How? Well thanks to Flutter GPU! A Dart layer around the Impeller rendering pipeline! This is all possible thanks to the awesome work from Brandon DeRosier 1/3

Flutter 💙 + Flame 🔥 = 3D?? What you see in the video is rendered by Flutter, no native packages just Flutter & Flame! How? Well thanks to Flutter GPU! A Dart layer around the Impeller rendering pipeline! This is all possible thanks to the awesome work from Brandon DeRosier 1/3

wolfenrain 💙🏳️‍🌈

21,782 次观看 • 2 年前

You may have seen this if you were at flutterCon last week, but this endless runner is a vanilla Flutter app. :) Dashsurfers (...name subject to change) is a little 3D game built with Flutter Scene, running in real time at 120 fps. The rendering, the particles, the UI, all Flutter.

You may have seen this if you were at flutterCon last week, but this endless runner is a vanilla Flutter app. :) Dashsurfers (...name subject to change) is a little 3D game built with Flutter Scene, running in real time at 120 fps. The rendering, the particles, the UI, all Flutter.

Brandon DeRosier

15,167 次观看 • 8 天前

Everything in this video is real-time 3D rendered in Flutter. No forks, no platform views, no secondary graphics context. Rendering, physics, gaussian splats, DICOM volumes, accessible scenes, embedded widgets, hot reload for every asset, instanced prefabs with arbitrary overrides, and much more. Get started at

Everything in this video is real-time 3D rendered in Flutter. No forks, no platform views, no secondary graphics context. Rendering, physics, gaussian splats, DICOM volumes, accessible scenes, embedded widgets, hot reload for every asset, instanced prefabs with arbitrary overrides, and much more. Get started at

Brandon DeRosier

17,293 次观看 • 10 天前

Tesla has integrated Unreal Engine into its vehicle software for rendering 3D car models. When zoomed in, it literally looks like a video game. The detail is incredible.

Tesla has integrated Unreal Engine into its vehicle software for rendering 3D car models. When zoomed in, it literally looks like a video game. The detail is incredible.

Nic Cruz Patane

109,821 次观看 • 2 个月前

The latest free version of UE5 AVATAR is now available! This update enhances real-time rendering quality and adds hundreds of new digital clothing and accessories. Create personalized 3D models, customize outfits, select scenes, and connect AI motion capture for real-time rendered 3D animations. Anyone can create 3D character models—no professional modeling skills required. Future versions will support 3D model import/export and image-based 3D model generation. Get the download link in the comments!

CYANPUPPETS

16,823 次观看 • 1 年前

The Future of Rendering keynote at #RenderCon2026 with Jules Urbach and Ari Emanuel is now available to watch. Here's everything covered: 00:46 - The state of rendering in 2026 01:36 - MCP for rendering 04:06 - Neural rendering and volumetric media 06:33 - World models 09:11 - Genie 3 demonstrations 17:22 - Creative workflows with generative 3D models 24:24 - Relightable Gaussian splats and virtual production 28:39 - Gene Roddenberry Archive and 3D Star Trek 32:35 - Jules Urbach and Ari Emanuel kick off a fireside chat 34:53 - Content production and cost reduction 38:41 - 3D models in sports and gaming 42:02 - AI's impact on storytelling and creative roles 45:17 - IP and actor digital rights Watch the full talk:

The Future of Rendering keynote at #RenderCon2026 with Jules Urbach and Ari Emanuel is now available to watch. Here's everything covered: 00:46 - The state of rendering in 2026 01:36 - MCP for rendering 04:06 - Neural rendering and volumetric media 06:33 - World models 09:11 - Genie 3 demonstrations 17:22 - Creative workflows with generative 3D models 24:24 - Relightable Gaussian splats and virtual production 28:39 - Gene Roddenberry Archive and 3D Star Trek 32:35 - Jules Urbach and Ari Emanuel kick off a fireside chat 34:53 - Content production and cost reduction 38:41 - 3D models in sports and gaming 42:02 - AI's impact on storytelling and creative roles 45:17 - IP and actor digital rights Watch the full talk:

The Render Network

21,948 次观看 • 2 个月前

🍺 LagerNVS (CVPR 2026) 🍺 LagerNVS is a generalizable, feed-forward, real-time Novel View Synthesis network which - performs rendering in real time, - generalizes to in-the-wild data, - works with and without known source cameras, - sets a new state-of-the-art among deterministic methods, - can be paired with a diffusion decoder for generative extrapolation. LagerNVS shows that 3D biases are useful for Novel View Synthesis but explicit 3D representations are not required to achieve them. We use 3D biases in (1) architecture design and (2) pre-training: (1) In NVS with explicit 3D representations (3DGS, NeRF) reconstruction is typically difficult and slow, but rendering is much faster and simpler. We mimic this process in the network design: we use a large (1B params) encoder and a small, lightweight decoder (ViT-B). This allows increasing the network capacity while still achieving real-time rendering. (2) The encoder, initialized from VGGT, was pre-trained with 3D reconstruction objectives, making the initial features 3D aware. Both substantially improve performance. Project page: Code: Paper: Models: Work done with Jianyuan Minghao Chen Christian Rupprecht and Andrea Vedaldi

🍺 LagerNVS (CVPR 2026) 🍺 LagerNVS is a generalizable, feed-forward, real-time Novel View Synthesis network which - performs rendering in real time, - generalizes to in-the-wild data, - works with and without known source cameras, - sets a new state-of-the-art among deterministic methods, - can be paired with a diffusion decoder for generative extrapolation. LagerNVS shows that 3D biases are useful for Novel View Synthesis but explicit 3D representations are not required to achieve them. We use 3D biases in (1) architecture design and (2) pre-training: (1) In NVS with explicit 3D representations (3DGS, NeRF) reconstruction is typically difficult and slow, but rendering is much faster and simpler. We mimic this process in the network design: we use a large (1B params) encoder and a small, lightweight decoder (ViT-B). This allows increasing the network capacity while still achieving real-time rendering. (2) The encoder, initialized from VGGT, was pre-trained with 3D reconstruction objectives, making the initial features 3D aware. Both substantially improve performance. Project page: Code: Paper: Models: Work done with Jianyuan Minghao Chen Christian Rupprecht and Andrea Vedaldi

Stan Szymanowicz

31,651 次观看 • 4 个月前

yall wanna see an up close shot of the face rigging work ive done for bbno$'s 3D model.. arkit + some correctives for better tongue and cheek movement. demo recorded in blender

yall wanna see an up close shot of the face rigging work ive done for bbno$'s 3D model.. arkit + some correctives for better tongue and cheek movement. demo recorded in blender

MADS ☀💙

47,586 次观看 • 9 个月前

Im decoupling rendering part of this project because WebGPU is so fresh and needs lots of work before it starts to look like I want. Good PBR rendering is super critical for app like this, game hairstyles should be authored in proper light environment. Trying to combine stylized anisotropic reflections with IBL lighting, fed fable 5 some research papers so it can cook better three.js

Im decoupling rendering part of this project because WebGPU is so fresh and needs lots of work before it starts to look like I want. Good PBR rendering is super critical for app like this, game hairstyles should be authored in proper light environment. Trying to combine stylized anisotropic reflections with IBL lighting, fed fable 5 some research papers so it can cook better three.js

Vadim B

10,973 次观看 • 29 天前

🚀3D GenAI is getting much better! 🤔3D generation now matches 3D printing in geometry, but color quantization remains a big issue: printers struggle with complex colors, and generators struggle with pure colors. 💪Rodin will solve this in the future.

🚀3D GenAI is getting much better! 🤔3D generation now matches 3D printing in geometry, but color quantization remains a big issue: printers struggle with complex colors, and generators struggle with pure colors. 💪Rodin will solve this in the future.

Hyper3D 🔜 GDC

94,154 次观看 • 8 个月前

We are looking at the future of VR, YouTube & Google Street View. This is zip-NeRF, a 3D neural rendering tech rapidly approaching the quality of a real, high-res drone flight video. Think of NeRF as transporting reality into simulation. Metaverse will finally work this time.

We are looking at the future of VR, YouTube & Google Street View. This is zip-NeRF, a 3D neural rendering tech rapidly approaching the quality of a real, high-res drone flight video. Think of NeRF as transporting reality into simulation. Metaverse will finally work this time.

Jim Fan

1,260,433 次观看 • 3 年前

I reverse engineered two more drone apps, adding support for 20 new drone models in TurboDrone including my new favorite the M10 which has brushless motors. Video feeds on this implementation use RTSP which is super smooth and reliable compared to the UDP sliced up JPEG frames I’ve had to deal with in the past. Cheap WiFi camera drones just keep getting better and better.

I reverse engineered two more drone apps, adding support for 20 new drone models in TurboDrone including my new favorite the M10 which has brushless motors. Video feeds on this implementation use RTSP which is super smooth and reliable compared to the UDP sliced up JPEG frames I’ve had to deal with in the past. Cheap WiFi camera drones just keep getting better and better.

Marshall Richards

45,760 次观看 • 3 个月前

Missed the NVIDIA keynote at #SIGGRAPH2026? Catch up on the latest breakthroughs in neural rendering, world models and simulation for robotics, presented by NVIDIA research and engineering leaders Neil Ashton, Edward Liu and Ming-Yu Liu. Watch on demand 🎥

Missed the NVIDIA keynote at #SIGGRAPH2026? Catch up on the latest breakthroughs in neural rendering, world models and simulation for robotics, presented by NVIDIA research and engineering leaders Neil Ashton, Edward Liu and Ming-Yu Liu. Watch on demand 🎥

NVIDIA

93,225 次观看 • 10 天前

This level of 3D tracking is INSANE. + Period Pain Simulator in full body~ I felt like showcasing how props can work paired with some effects all done live and handled by 1 person! ( Bonus, you get to see what the gloves and prop looks like )

This level of 3D tracking is INSANE. + Period Pain Simulator in full body~ I felt like showcasing how props can work paired with some effects all done live and handled by 1 person! ( Bonus, you get to see what the gloves and prop looks like )

Azeru Official

58,603 次观看 • 11 个月前

This is what GOALS looks like in open alpha, and it's only getting better. Come play and help us shape the game!

This is what GOALS looks like in open alpha, and it's only getting better. Come play and help us shape the game!

GOALS

12,177 次观看 • 1 年前

Decentralized Rendering hitting home runs 🏈 ! The LA Rams’ Los Angeles Rams 4th quarter hype video, premiered directly from the SoFi Stadium Infinity Screen on November 3rd, was fully rendered on OctaRender our decentralized rendering platform. The scene was shaped for a 15:1 aspect ratio, wrapping around a 360° stadium screen and rendered in 8K and upscaled to 15K resolution. Directed by Flint Tanquary VFX Lead - Ezekiel Wagner 3D Generalists - Isfahan Ahmed, Numaan Shaikh

Decentralized Rendering hitting home runs 🏈 ! The LA Rams’ Los Angeles Rams 4th quarter hype video, premiered directly from the SoFi Stadium Infinity Screen on November 3rd, was fully rendered on OctaRender our decentralized rendering platform. The scene was shaped for a 15:1 aspect ratio, wrapping around a 360° stadium screen and rendered in 8K and upscaled to 15K resolution. Directed by Flint Tanquary VFX Lead - Ezekiel Wagner 3D Generalists - Isfahan Ahmed, Numaan Shaikh

OctaSpace

37,936 次观看 • 8 个月前

AI fully taking over 3D is inevitable you can now add 3D bg for 3D models inside 3d software using World Labs's Marble and.. view in real time and even use Octane's 3DGS to light up the model without adding lights this is impossible w/o AI

AI fully taking over 3D is inevitable you can now add 3D bg for 3D models inside 3d software using World Labs's Marble and.. view in real time and even use Octane's 3DGS to light up the model without adding lights this is impossible w/o AI

el.cine

174,533 次观看 • 8 个月前

GaVS: 3D-Grounded Video Stabilization via Temporally-Consistent Local Reconstruction and Rendering Contributions: • We reformulate video stabilization as a novel 3D grounded scheme of local reconstruction and rendering. This approach is naturally robust to diverse camera motions and scene dynamics, is temporally consistent, and is capable of full frame stabilization. • We propose a novel test-time optimization for each unstable video. It leverages multi-view dynamics-aware photometric supervision and cross-frame regularization to achieve temporally consistent reconstructions. To avoid frame cropping, we introduce a scene extrapolation module based on video completion. • We provide a 3D-grounded dataset for our task by re-purposing an existing one, and introduce new metrics on sparse and dense reconstruction to evaluate 3D scene consistency. Extensive experiments (quantitative, qualitative, user study) versus image-based and gyro-basedmethods demonstrate the merits of our method.

GaVS: 3D-Grounded Video Stabilization via Temporally-Consistent Local Reconstruction and Rendering Contributions: • We reformulate video stabilization as a novel 3D grounded scheme of local reconstruction and rendering. This approach is naturally robust to diverse camera motions and scene dynamics, is temporally consistent, and is capable of full frame stabilization. • We propose a novel test-time optimization for each unstable video. It leverages multi-view dynamics-aware photometric supervision and cross-frame regularization to achieve temporally consistent reconstructions. To avoid frame cropping, we introduce a scene extrapolation module based on video completion. • We provide a 3D-grounded dataset for our task by re-purposing an existing one, and introduce new metrics on sparse and dense reconstruction to evaluate 3D scene consistency. Extensive experiments (quantitative, qualitative, user study) versus image-based and gyro-basedmethods demonstrate the merits of our method.

MrNeRF

11,638 次观看 • 1 年前

3D-LLM: Injecting the 3D World into Large Language Models paper page: Large language models (LLMs) and Vision-Language Models (VLMs) have been proven to excel at multiple tasks, such as commonsense reasoning. Powerful as these models can be, they are not grounded in the 3D physical world, which involves richer concepts such as spatial relationships, affordances, physics, layout, and so on. In this work, we propose to inject the 3D world into large language models and introduce a whole new family of 3D-LLMs. Specifically, 3D-LLMs can take 3D point clouds and their features as input and perform a diverse set of 3D-related tasks, including captioning, dense captioning, 3D question answering, task decomposition, 3D grounding, 3D-assisted dialog, navigation, and so on. Using three types of prompting mechanisms that we design, we are able to collect over 300k 3D-language data covering these tasks. To efficiently train 3D-LLMs, we first utilize a 3D feature extractor that obtains 3D features from rendered multi- view images. Then, we use 2D VLMs as our backbones to train our 3D-LLMs. By introducing a 3D localization mechanism, 3D-LLMs can better capture 3D spatial information. Experiments on ScanQA show that our model outperforms state-of-the-art baselines by a large margin (e.g., the BLEU-1 score surpasses state-of-the-art score by 9%). Furthermore, experiments on our held-in datasets for 3D captioning, task composition, and 3D-assisted dialogue show that our model outperforms 2D VLMs. Qualitative examples also show that our model could perform more tasks beyond the scope of existing LLMs and VLMs.

3D-LLM: Injecting the 3D World into Large Language Models paper page: Large language models (LLMs) and Vision-Language Models (VLMs) have been proven to excel at multiple tasks, such as commonsense reasoning. Powerful as these models can be, they are not grounded in the 3D physical world, which involves richer concepts such as spatial relationships, affordances, physics, layout, and so on. In this work, we propose to inject the 3D world into large language models and introduce a whole new family of 3D-LLMs. Specifically, 3D-LLMs can take 3D point clouds and their features as input and perform a diverse set of 3D-related tasks, including captioning, dense captioning, 3D question answering, task decomposition, 3D grounding, 3D-assisted dialog, navigation, and so on. Using three types of prompting mechanisms that we design, we are able to collect over 300k 3D-language data covering these tasks. To efficiently train 3D-LLMs, we first utilize a 3D feature extractor that obtains 3D features from rendered multi- view images. Then, we use 2D VLMs as our backbones to train our 3D-LLMs. By introducing a 3D localization mechanism, 3D-LLMs can better capture 3D spatial information. Experiments on ScanQA show that our model outperforms state-of-the-art baselines by a large margin (e.g., the BLEU-1 score surpasses state-of-the-art score by 9%). Furthermore, experiments on our held-in datasets for 3D captioning, task composition, and 3D-assisted dialogue show that our model outperforms 2D VLMs. Qualitative examples also show that our model could perform more tasks beyond the scope of existing LLMs and VLMs.

AK

249,708 次观看 • 3 年前

I'll tell it like it is: they're lying. Robots are getting better and better, and the goal now is not to support human labor, as suggested in the video, but in the long term they will replace human labor. In a market economy, profit is the only thing that counts, whether you like to criticize that or not. The fact is, however, that robots will be cheaper than human labor and will therefore be preferred. Moreover, robots are getting better and better, they will work longer hours than humans, are less prone to errors, do not fall ill and do not need vacations. They are superior to human labor.

I'll tell it like it is: they're lying. Robots are getting better and better, and the goal now is not to support human labor, as suggested in the video, but in the long term they will replace human labor. In a market economy, profit is the only thing that counts, whether you like to criticize that or not. The fact is, however, that robots will be cheaper than human labor and will therefore be preferred. Moreover, robots are getting better and better, they will work longer hours than humans, are less prone to errors, do not fall ill and do not need vacations. They are superior to human labor.

Chubby♨️

65,166 次观看 • 1 年前