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。

Come to our #SIGGRAPH2026 talk today: Our method turns existing 3D objects into multi-touch sensors, without modifying the interior. From a scan, we sample surface-intrinsic curves that conform to mutual-cap layouts, and optimize their intersections for coverage & fabrication.

Christian Holz

2,546 subscribers

52,181 次观看 • 12 天前 •via X (Twitter)

科学技术教育 #SIGGRAPH2026

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

Blended-NeRF: Zero-Shot Object Generation and Blending in Existing Neural Radiance Fields paper page: Editing a local region or a specific object in a 3D scene represented by a NeRF is challenging, mainly due to the implicit nature of the scene representation. Consistently blending a new realistic object into the scene adds an additional level of difficulty. We present Blended-NeRF, a robust and flexible framework for editing a specific region of interest in an existing NeRF scene, based on text prompts or image patches, along with a 3D ROI box. Our method leverages a pretrained language-image model to steer the synthesis towards a user-provided text prompt or image patch, along with a 3D MLP model initialized on an existing NeRF scene to generate the object and blend it into a specified region in the original scene. We allow local editing by localizing a 3D ROI box in the input scene, and seamlessly blend the content synthesized inside the ROI with the existing scene using a novel volumetric blending technique. To obtain natural looking and view-consistent results, we leverage existing and new geometric priors and 3D augmentations for improving the visual fidelity of the final result. We test our framework both qualitatively and quantitatively on a variety of real 3D scenes and text prompts, demonstrating realistic multi-view consistent results with much flexibility and diversity compared to the baselines. Finally, we show the applicability of our framework for several 3D editing applications, including adding new objects to a scene, removing/replacing/altering existing objects, and texture conversion.

Blended-NeRF: Zero-Shot Object Generation and Blending in Existing Neural Radiance Fields paper page: Editing a local region or a specific object in a 3D scene represented by a NeRF is challenging, mainly due to the implicit nature of the scene representation. Consistently blending a new realistic object into the scene adds an additional level of difficulty. We present Blended-NeRF, a robust and flexible framework for editing a specific region of interest in an existing NeRF scene, based on text prompts or image patches, along with a 3D ROI box. Our method leverages a pretrained language-image model to steer the synthesis towards a user-provided text prompt or image patch, along with a 3D MLP model initialized on an existing NeRF scene to generate the object and blend it into a specified region in the original scene. We allow local editing by localizing a 3D ROI box in the input scene, and seamlessly blend the content synthesized inside the ROI with the existing scene using a novel volumetric blending technique. To obtain natural looking and view-consistent results, we leverage existing and new geometric priors and 3D augmentations for improving the visual fidelity of the final result. We test our framework both qualitatively and quantitatively on a variety of real 3D scenes and text prompts, demonstrating realistic multi-view consistent results with much flexibility and diversity compared to the baselines. Finally, we show the applicability of our framework for several 3D editing applications, including adding new objects to a scene, removing/replacing/altering existing objects, and texture conversion.

AK

62,768 次观看 • 3 年前

Multi-view Reconstruction via SfM-guided Monocular Depth Estimation Contributions: • We propose a novel approach to inject SfM priors into diffusion-based depth estimation, enabling highly accurate and multi-view consistent depth predictions for each viewpoint. • Based on the proposed depth estimator, we design a new multi-view 3D geometry reconstruction framework and process some synthetic datasets to facilitate training. • We evaluate our method on diverse real-world scene data, including objects, indoor environments, streetscapes, and aerial scenes, demonstrating the superior performance and generalization capability of our approach.

Multi-view Reconstruction via SfM-guided Monocular Depth Estimation Contributions: • We propose a novel approach to inject SfM priors into diffusion-based depth estimation, enabling highly accurate and multi-view consistent depth predictions for each viewpoint. • Based on the proposed depth estimator, we design a new multi-view 3D geometry reconstruction framework and process some synthetic datasets to facilitate training. • We evaluate our method on diverse real-world scene data, including objects, indoor environments, streetscapes, and aerial scenes, demonstrating the superior performance and generalization capability of our approach.

MrNeRF

25,651 次观看 • 1 年前

To turn our movement into a vehicle for our ideas and our people, we need to first build that movement from the ground up

To turn our movement into a vehicle for our ideas and our people, we need to first build that movement from the ground up

Fuentes Updates

29,791 次观看 • 1 年前

Quebec is officially calling for an end to multiculturalism & requiring people who come into the province to conform. His plan makes sense. It allows people to retain their culture, but conform to our values If I were a one-issue voter, this would hit top three for me. All day

Quebec is officially calling for an end to multiculturalism & requiring people who come into the province to conform. His plan makes sense. It allows people to retain their culture, but conform to our values If I were a one-issue voter, this would hit top three for me. All day

Mario Zelaya

34,435 次观看 • 1 年前

Our TEDxLondon talk is live for #IWD2023 and we want to talk about feminist liberation & care! Collective care is our strongest tool to build a world without violence. Here’s a short 🧵 on why & you can watch the full talk here:

Our TEDxLondon talk is live for #IWD2023 and we want to talk about feminist liberation & care! Collective care is our strongest tool to build a world without violence. Here’s a short 🧵 on why & you can watch the full talk here:

Level Up

32,254 次观看 • 3 年前

Today, we announced a partnership with some of tech’s most recognizable names to develop new ways to optimize our grid for customers. We’re applying expertise from @IonQ, Oak Ridge Lab and @NVIDIA to create new solutions and prepare our community for the future.

Today, we announced a partnership with some of tech’s most recognizable names to develop new ways to optimize our grid for customers. We’re applying expertise from @IonQ, Oak Ridge Lab and @NVIDIA to create new solutions and prepare our community for the future.

EPB

58,623 次观看 • 10 个月前

Multi-Agent workflows are the future of AI. OpenAI released new Agent APIs today, and Box built an Agent that combines documents from Box and web search tools to generate answers. Enterprise devs can grab sample code from our GitHub repo to customize with their data.

Multi-Agent workflows are the future of AI. OpenAI released new Agent APIs today, and Box built an Agent that combines documents from Box and web search tools to generate answers. Enterprise devs can grab sample code from our GitHub repo to customize with their data.

Aaron Levie

102,062 次观看 • 1 年前

Today, we honor the story of our veterans – the story of our nation at its best. On Veterans Day, let’s recommit to fulfilling our one sacred obligation as a nation: to prepare those we send into harm’s way and care for them and their families when they come home.

Today, we honor the story of our veterans – the story of our nation at its best. On Veterans Day, let’s recommit to fulfilling our one sacred obligation as a nation: to prepare those we send into harm’s way and care for them and their families when they come home.

President Biden Archived

1,803,180 次观看 • 2 年前

Talk to me about compassion and humanity for undocumented immigrants. America is the leader of compassion for citizens, No matter where they come from. Not only do we grant them benefits that are exclusively for American citizens during their illegal stay We also provide $1000 If self-deport. And the opportunity to apply for legal entry into this country. See in comparison the treatment that the brother" country of Mexico gives to American citizens when without any bad intention But without permission from the immigration authorities. We extend our tourist visa⬇️ #LosAngelesRiots

Talk to me about compassion and humanity for undocumented immigrants. America is the leader of compassion for citizens, No matter where they come from. Not only do we grant them benefits that are exclusively for American citizens during their illegal stay We also provide $1000 If self-deport. And the opportunity to apply for legal entry into this country. See in comparison the treatment that the brother" country of Mexico gives to American citizens when without any bad intention But without permission from the immigration authorities. We extend our tourist visa⬇️ #LosAngelesRiots

Real Hamlet Garcia

33,374 次观看 • 1 年前

Europe is where we come from. And it’s where we are headed. Because we are Europeans. Today, I addressed the European Parliament on our commitment to a European path. Without Belarus and Ukraine, the EU remains incomplete. This is the future we passionately work towards for our people.

Europe is where we come from. And it’s where we are headed. Because we are Europeans. Today, I addressed the European Parliament on our commitment to a European path. Without Belarus and Ukraine, the EU remains incomplete. This is the future we passionately work towards for our people.

Sviatlana Tsikhanouskaya

54,698 次观看 • 2 年前

"More Doggo than Doggo" Motion scanned @ 60fps. Playing back @ 20fps. Meet our inhouse tech Doggo "Rolo". Since July we've been redesigning our scanning pipeline to work with the new 3D Gaussian Splatting for Real-Time Radiance Field Rendering method from Inria. GS has reignited our passion for 3D/4D scanning. Thank you to the team at Inria IR's AeonX capture system has been designed to capture "Spatial" memories for the future. We've built on a custom version of Inria's Sibr viewer to handle the playback of our multiple processed 4D frames. Our next goal is to optimize the player to be multi-threaded to improve performance over 60fps. GS has reignited our passion for 3D/4D scanning. Thank you to the team at Inria More content to come. #guassiansplatting #ir #aeonx #4d #motionscanning #inria #ximea #idatronic #spatial #youth83

"More Doggo than Doggo" Motion scanned @ 60fps. Playing back @ 20fps. Meet our inhouse tech Doggo "Rolo". Since July we've been redesigning our scanning pipeline to work with the new 3D Gaussian Splatting for Real-Time Radiance Field Rendering method from Inria. GS has reignited our passion for 3D/4D scanning. Thank you to the team at Inria IR's AeonX capture system has been designed to capture "Spatial" memories for the future. We've built on a custom version of Inria's Sibr viewer to handle the playback of our multiple processed 4D frames. Our next goal is to optimize the player to be multi-threaded to improve performance over 60fps. GS has reignited our passion for 3D/4D scanning. Thank you to the team at Inria More content to come. #guassiansplatting #ir #aeonx #4d #motionscanning #inria #ximea #idatronic #spatial #youth83

Infinite-Realities

48,329 次观看 • 2 年前

It’s always cool to see an idea come to life. We started our Houston Rockets broadcast with a walk-and-talk that we recorded as Jabari entered the arena! Thank you Jabari Smith Jr for your time and insight & our Space City Home Network crew for the hard work.

It’s always cool to see an idea come to life. We started our Houston Rockets broadcast with a walk-and-talk that we recorded as Jabari entered the arena! Thank you Jabari Smith Jr for your time and insight & our Space City Home Network crew for the hard work.

Vanessa Richardson

38,791 次观看 • 8 个月前

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

The problems we're facing in our country today stem from one simple reason: Our generation doesn't know how to talk. We were raised to believe that politics and faith were taboo topics. That was a mistake. We need to bring back civil conversations.

The problems we're facing in our country today stem from one simple reason: Our generation doesn't know how to talk. We were raised to believe that politics and faith were taboo topics. That was a mistake. We need to bring back civil conversations.

Brilyn Hollyhand

44,331 次观看 • 9 个月前

Spatial reasoning is a major challenge for the foundation models today, even in simple tasks like arranging objects in 3D space. #CVPR2025 Introducing LayoutVLM, a differentiable optimization framework that uses VLM to spatially reason about diverse scene layouts from unlabeled assets and open-ended language instructions 1/n

Spatial reasoning is a major challenge for the foundation models today, even in simple tasks like arranging objects in 3D space. #CVPR2025 Introducing LayoutVLM, a differentiable optimization framework that uses VLM to spatially reason about diverse scene layouts from unlabeled assets and open-ended language instructions 1/n

Fan-Yun Sun

92,584 次观看 • 1 年前

Grateful to the Legal Defense Fund and grateful to be in Selma today to talk about this moment. This is our time and our opportunity to build a country that those who came before us hoped for — and that those who come after us deserve.

Grateful to the Legal Defense Fund and grateful to be in Selma today to talk about this moment. This is our time and our opportunity to build a country that those who came before us hoped for — and that those who come after us deserve.

Wes Moore

13,757 次观看 • 4 个月前

NEW: Robert F. Kennedy Jr Officially Announces His Run for the Presidency as an Independent "I'm here to join you in making a new declaration of independence for our entire nation. We declare independence from the corporations that have hijacked our government. And we declare independence from Wall Street, from Big Tech, from Big Pharma, from Big AG, from the military contractors and their lobbyists. And we declare independence from the mercenary media that is here to fortify all of the corporate orthodoxies from their advertisers and to urge us to hate our neighbors and to fear our friends. And we declare independence from the cynical elites who betray our hope and who amplify our divisions."

NEW: Robert F. Kennedy Jr Officially Announces His Run for the Presidency as an Independent "I'm here to join you in making a new declaration of independence for our entire nation. We declare independence from the corporations that have hijacked our government. And we declare independence from Wall Street, from Big Tech, from Big Pharma, from Big AG, from the military contractors and their lobbyists. And we declare independence from the mercenary media that is here to fortify all of the corporate orthodoxies from their advertisers and to urge us to hate our neighbors and to fear our friends. And we declare independence from the cynical elites who betray our hope and who amplify our divisions."

The Vigilant Fox 🦊

500,785 次观看 • 2 年前

SceNeRFlow: Time-Consistent Reconstruction of General Dynamic Scenes abs: paper page: Existing methods for the 4D reconstruction of general, non-rigidly deforming objects focus on novel-view synthesis and neglect correspondences. However, time consistency enables advanced downstream tasks like 3D editing, motion analysis, or virtual-asset creation. We propose SceNeRFlow to reconstruct a general, non-rigid scene in a time-consistent manner. Our dynamic-NeRF method takes multi-view RGB videos and background images from static cameras with known camera parameters as input. It then reconstructs the deformations of an estimated canonical model of the geometry and appearance in an online fashion. Since this canonical model is time-invariant, we obtain correspondences even for long-term, long-range motions. We employ neural scene representations to parametrize the components of our method. Like prior dynamic-NeRF methods, we use a backwards deformation model. We find non-trivial adaptations of this model necessary to handle larger motions: We decompose the deformations into a strongly regularized coarse component and a weakly regularized fine component, where the coarse component also extends the deformation field into the space surrounding the object, which enables tracking over time. We show experimentally that, unlike prior work that only handles small motion, our method enables the reconstruction of studio-scale motions.

SceNeRFlow: Time-Consistent Reconstruction of General Dynamic Scenes abs: paper page: Existing methods for the 4D reconstruction of general, non-rigidly deforming objects focus on novel-view synthesis and neglect correspondences. However, time consistency enables advanced downstream tasks like 3D editing, motion analysis, or virtual-asset creation. We propose SceNeRFlow to reconstruct a general, non-rigid scene in a time-consistent manner. Our dynamic-NeRF method takes multi-view RGB videos and background images from static cameras with known camera parameters as input. It then reconstructs the deformations of an estimated canonical model of the geometry and appearance in an online fashion. Since this canonical model is time-invariant, we obtain correspondences even for long-term, long-range motions. We employ neural scene representations to parametrize the components of our method. Like prior dynamic-NeRF methods, we use a backwards deformation model. We find non-trivial adaptations of this model necessary to handle larger motions: We decompose the deformations into a strongly regularized coarse component and a weakly regularized fine component, where the coarse component also extends the deformation field into the space surrounding the object, which enables tracking over time. We show experimentally that, unlike prior work that only handles small motion, our method enables the reconstruction of studio-scale motions.

AK

76,380 次观看 • 3 年前

Omkara turns 2 years old today. From 2 employees when we started to over 30 employees today, and from searching for our first client to over 2,000 happy clients today, it has been an incredible journey. A big thank you to everyone for the love, support, trust & criticism. ❤️se 🙏

Omkara turns 2 years old today. From 2 employees when we started to over 30 employees today, and from searching for our first client to over 2,000 happy clients today, it has been an incredible journey. A big thank you to everyone for the love, support, trust & criticism. ❤️se 🙏

Varinder Bansal 🇮🇳

201,082 次观看 • 3 年前

Today marks the first day without federal funding for public media — a serious setback for the entire PBS system. Even so, our commitment to you hasn’t wavered. We remain steadfast in our responsibility to deliver reporting that is fair, accurate and independent. And, we'll continue to provide the kind of trusted, in-depth coverage that has defined public media for generations. For nearly 50 years, PBS News and this nightly program have given you the facts, context and insight to understand our world and make informed decisions about your lives. That mission drives our work every day. Your support makes that possible, and for that, we are grateful.

Today marks the first day without federal funding for public media — a serious setback for the entire PBS system. Even so, our commitment to you hasn’t wavered. We remain steadfast in our responsibility to deliver reporting that is fair, accurate and independent. And, we'll continue to provide the kind of trusted, in-depth coverage that has defined public media for generations. For nearly 50 years, PBS News and this nightly program have given you the facts, context and insight to understand our world and make informed decisions about your lives. That mission drives our work every day. Your support makes that possible, and for that, we are grateful.

PBS News

21,198 次观看 • 10 个月前