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。

🚨 ConceptGraphs: Open-Vocabulary 3D Scene Graphs for Perception and Planning RGB-D images -> 3D scene-graphs -> enhance with vision-language features Enables a wide range of perception and planning abilities. Deployed in many real-world use-cases.

Krishna Murthy

4,776 subscribers

75,593 次观看 • 2 年前 •via X (Twitter)

科学技术教育

Anya Rossi• Live Now

Private livecam show

11 条评论

Krishna Murthy 的头像

Krishna Murthy2 年前

We designed a 3D mapping pipeline that takes in posed RGB-D images and builds an object-based map. Each object in the map is augmented with multi-view fused CLIP features and language descriptions. We build a scene graph; which can easily be parsed by LLMs

Krishna Murthy 的头像

Krishna Murthy2 年前

Text queries can be abstract / complex. We ask our robot to find "something that goes well with a Ronald MacDonald outfit", and it finds a pair of "red and white sneakers"

Krishna Murthy 的头像

Krishna Murthy2 年前

We can also identify misplaced objects in the scene graph and search for them. Here, we move the shoes away from the earlier location. The robot then begins searching for plausible locations for the red-white sneakers, and finds them at a nearby shoerack

Krishna Murthy 的头像

Krishna Murthy2 年前

Another query: "Something to wear for a space party" Finds a t-shirt that has NASA written on it

Krishna Murthy 的头像

Krishna Murthy2 年前

Queries can involve both text and image context. Show the robot a photograph of Michael Jordan, and ask it to go to "something this guy would play with" The robot finds a nearby basketball

Krishna Murthy 的头像

Krishna Murthy2 年前

Another example. Here, a handwritten note that reads "go to the laundry bag" indicates to the robot its target location.

Krishna Murthy 的头像

Krishna Murthy2 年前

This time, a mobile manipulation task, where a Spot mini has to pick up "something healthy to eat". It picks up a plastic mango

Krishna Murthy 的头像

Krishna Murthy2 年前

"My wrist hurts from using my screwdriver all day. Anything to help?" Robot finds a power drill

Krishna Murthy 的头像

Krishna Murthy2 年前

We implement a particle-filter based localization approach using per-object CLIP features. This helps us relocalize against a prebuilt map; identify and add new objects to the map; and detect objects that no longer exist in the scene.

Krishna Murthy 的头像

Krishna Murthy2 年前

We're excited about the prospects ConceptGraphs brings about! Great fun collab: HUGE shoutout to @qiaogu1997 @alihkw_ @SachaMori without whom this work wouldn't have taken the shape it has w/ @bipashasen31 @skymanaditya1 @duckietown_coo @florian_shkurti and many others

Krishna Murthy 的头像

Krishna Murthy2 年前

ConceptGraphs significantly improves upon ConceptFusion ( making it possible to scale to larger scenes, run inference on-board CPUs, and handles more complex queries. Find Ali's short explainer video at:

相关视频

LEGO-SLAM: Language-Embedded Gaussian Optimization SLAM LEGO-SLAM running at 15 FPS on a ScanNet scene with language-based loop closing for drift correction. LEGO-SLAM is a 3DGS-based SLAM framework that supports open-vocabulary semantic querying and rendering. It tracks via G-ICP and efficiently builds a map by embedding Gaussians with scene-adaptive 16D language features. Map management is achieved through Language Pruning and Language-Based Loop Detection. The generated map enables open-vocabulary 3D Object Localization.

LEGO-SLAM: Language-Embedded Gaussian Optimization SLAM LEGO-SLAM running at 15 FPS on a ScanNet scene with language-based loop closing for drift correction. LEGO-SLAM is a 3DGS-based SLAM framework that supports open-vocabulary semantic querying and rendering. It tracks via G-ICP and efficiently builds a map by embedding Gaussians with scene-adaptive 16D language features. Map management is achieved through Language Pruning and Language-Based Loop Detection. The generated map enables open-vocabulary 3D Object Localization.

Ryohei Sasaki@engineer

14,935 次观看 • 3 个月前

📢ProcGen3D: Learning Neural Procedural Graphs for Image-to-3D Reconstruction Xinyi Zhang learns neural procedural graphs to generate high-fidelity 3D - MCTS-guided sampling maintains consistency with the input image, even from real images! Check it out:

📢ProcGen3D: Learning Neural Procedural Graphs for Image-to-3D Reconstruction Xinyi Zhang learns neural procedural graphs to generate high-fidelity 3D - MCTS-guided sampling maintains consistency with the input image, even from real images! Check it out:

Angela Dai

14,590 次观看 • 7 个月前

Chain-of-thought reasoning is a powerful tool to enable language models to work through complex problems. Can we use this with robots? With embodied chain-of-thought, vision-language-action (VLA) models can think through perception and planning! A 🧵👇

Chain-of-thought reasoning is a powerful tool to enable language models to work through complex problems. Can we use this with robots? With embodied chain-of-thought, vision-language-action (VLA) models can think through perception and planning! A 🧵👇

Sergey Levine

30,388 次观看 • 1 年前

LLaVA-3D A Simple yet Effective Pathway to Empowering LMMs with 3D-awareness Recent advancements in Large Multimodal Models (LMMs) have greatly enhanced their proficiency in 2D visual understanding tasks, enabling them to effectively process and understand images and videos. However, the development of LMMs with 3D-awareness for 3D scene understanding has been hindered by the lack of large-scale 3D vision-language datasets and powerful 3D encoders. In this paper, we introduce a simple yet effective framework called LLaVA-3D. Leveraging the strong 2D understanding priors from LLaVA, our LLaVA-3D efficiently adapts LLaVA for 3D scene understanding without compromising 2D understanding capabilities. To achieve this, we employ a simple yet effective representation, 3D Patch, which connects 2D CLIP patch features with their corresponding positions in 3D space. By integrating the 3D Patches into 2D LMMs and employing joint 2D and 3D vision-language instruction tuning, we establish a unified architecture for both 2D image understanding and 3D scene understanding. Experimental results show that LLaVA-3D converges 3.5x faster than existing 3D LMMs when trained on 3D vision-language datasets. Moreover, LLaVA-3D not only achieves state-of-the-art performance across various 3D tasks but also maintains comparable 2D image understanding and vision-language conversation capabilities with LLaVA.

LLaVA-3D A Simple yet Effective Pathway to Empowering LMMs with 3D-awareness Recent advancements in Large Multimodal Models (LMMs) have greatly enhanced their proficiency in 2D visual understanding tasks, enabling them to effectively process and understand images and videos. However, the development of LMMs with 3D-awareness for 3D scene understanding has been hindered by the lack of large-scale 3D vision-language datasets and powerful 3D encoders. In this paper, we introduce a simple yet effective framework called LLaVA-3D. Leveraging the strong 2D understanding priors from LLaVA, our LLaVA-3D efficiently adapts LLaVA for 3D scene understanding without compromising 2D understanding capabilities. To achieve this, we employ a simple yet effective representation, 3D Patch, which connects 2D CLIP patch features with their corresponding positions in 3D space. By integrating the 3D Patches into 2D LMMs and employing joint 2D and 3D vision-language instruction tuning, we establish a unified architecture for both 2D image understanding and 3D scene understanding. Experimental results show that LLaVA-3D converges 3.5x faster than existing 3D LMMs when trained on 3D vision-language datasets. Moreover, LLaVA-3D not only achieves state-of-the-art performance across various 3D tasks but also maintains comparable 2D image understanding and vision-language conversation capabilities with LLaVA.

AK

41,713 次观看 • 1 年前

Everyone is scaling VLAs with more robot data. TiPToP shows another path. No robot training, no policy learning. Just RGB + language → 3D scene → GPU TAMP planner → trajectory. Foundation models + planning alone can run real manipulation tasks.

Everyone is scaling VLAs with more robot data. TiPToP shows another path. No robot training, no policy learning. Just RGB + language → 3D scene → GPU TAMP planner → trajectory. Foundation models + planning alone can run real manipulation tasks.

Robots Digest 🤖

10,362 次观看 • 3 个月前

New video of AGIBOT's Lingxi X2 robot showcasing dynamic movements, fall recovery, quiet operation, and vision-based perception/planning.

New video of AGIBOT's Lingxi X2 robot showcasing dynamic movements, fall recovery, quiet operation, and vision-based perception/planning.

The Humanoid Hub

65,880 次观看 • 1 年前

SuperGSeg: Open-Vocabulary 3D Segmentation with Structured Super-Gaussians Contributions: • We propose SuperGSeg: a 3D segmentation method with neural Gaussians, designed to learn hierarchical instance segmentation features from 2D foundation models. • We introduce the concept of Super-Gaussian, a novel representation that integrates hierarchical instance segmentation features, enabling the embedding of high-dimensional language features. This approach addresses previously unfeasible challenges in representing complex scenes with rich semantic details. • Extensive experiments on the LERF-OVS and ScanNet datasets demonstrate the effectiveness of the proposed method, achieving significant improvements in open-vocabulary 3D object-level and scene-level semantic segmentation. It shows particular strength in capturing fine-grained scene details and dense pixel semantic segmentation tasks for the first time.

SuperGSeg: Open-Vocabulary 3D Segmentation with Structured Super-Gaussians Contributions: • We propose SuperGSeg: a 3D segmentation method with neural Gaussians, designed to learn hierarchical instance segmentation features from 2D foundation models. • We introduce the concept of Super-Gaussian, a novel representation that integrates hierarchical instance segmentation features, enabling the embedding of high-dimensional language features. This approach addresses previously unfeasible challenges in representing complex scenes with rich semantic details. • Extensive experiments on the LERF-OVS and ScanNet datasets demonstrate the effectiveness of the proposed method, achieving significant improvements in open-vocabulary 3D object-level and scene-level semantic segmentation. It shows particular strength in capturing fine-grained scene details and dense pixel semantic segmentation tasks for the first time.

MrNeRF

13,594 次观看 • 1 年前

Voyager Long-Range and World-Consistent Video Diffusion for Explorable 3D Scene Generation

Voyager Long-Range and World-Consistent Video Diffusion for Explorable 3D Scene Generation

AK

15,840 次观看 • 1 年前

📢Panoptic Lifting for 3D Scene Understanding with Neural Fields #CVPR23 highlight! Given only posed RGB images of a scene, we optimize a panoptic radiance field representing color, depth, semantics, and instances at any point in space. Vid: Yawar Siddiqui

📢Panoptic Lifting for 3D Scene Understanding with Neural Fields #CVPR23 highlight! Given only posed RGB images of a scene, we optimize a panoptic radiance field representing color, depth, semantics, and instances at any point in space. Vid: Yawar Siddiqui

Matthias Niessner

29,830 次观看 • 3 年前

$Voronoi diagram (black points) computed by projecting vertically lower envelope of n 3D graphs of functions {(x,y_i(x))} with y_i(x)=D(x_i,x) (pink). When distance D(x,x')=‖x-x'‖^2, graphs of y_i are paraboloids and Voronoi cell borders are linear$

Voronoi diagram (black points) computed by projecting vertically lower envelope of n 3D graphs of functions {(x,y_i(x))} with y_i(x)=D(x_i,x) (pink). When distance D(x,x')=‖x-x'‖^2, graphs of y_i are paraboloids and Voronoi cell borders are linear

Frank Nielsen

32,345 次观看 • 9 天前

$Voronoi diagram (black points) computed by projecting vertically lower envelope of n 3D graphs of functions {(x,y_i(x))} with y_i(x)=D(x_i,x) (pink). When distance D(x,x')=‖x-x'‖^2, graphs of y_i are paraboloids and Voronoi cell borders are linear$

Voronoi diagram (black points) computed by projecting vertically lower envelope of n 3D graphs of functions {(x,y_i(x))} with y_i(x)=D(x_i,x) (pink). When distance D(x,x')=‖x-x'‖^2, graphs of y_i are paraboloids and Voronoi cell borders are linear

Frank Nielsen

98,601 次观看 • 7 个月前

LangSplatV2: High-dimensional 3D Language Gaussian Splatting with 450+ FPS Contributions: • LangSplatV2 achieves real-time performance with 476.2 FPS for high-dimensional feature splatting and 384.6 FPS for 3D open-vocabulary text querying. • Delivers a 42× speedup and 47× boost in performance compared to LangSplat. • Improves query accuracy while drastically reducing inference time. • Replaces the heavyweight decoder in LangSplat with a sparse coefficient field, removing the main performance bottleneck. • Introduces a CUDA-optimized sparse coefficient splatting method, enabling fast and high-quality rendering of high-dimensional features. • Enables scalable 3D language interaction in complex scenes, opening up real-time applications previously not possible with LangSplat.

LangSplatV2: High-dimensional 3D Language Gaussian Splatting with 450+ FPS Contributions: • LangSplatV2 achieves real-time performance with 476.2 FPS for high-dimensional feature splatting and 384.6 FPS for 3D open-vocabulary text querying. • Delivers a 42× speedup and 47× boost in performance compared to LangSplat. • Improves query accuracy while drastically reducing inference time. • Replaces the heavyweight decoder in LangSplat with a sparse coefficient field, removing the main performance bottleneck. • Introduces a CUDA-optimized sparse coefficient splatting method, enabling fast and high-quality rendering of high-dimensional features. • Enables scalable 3D language interaction in complex scenes, opening up real-time applications previously not possible with LangSplat.

MrNeRF

14,261 次观看 • 11 个月前

Planning with Reasoning using Vision Language World Model

Planning with Reasoning using Vision Language World Model

AK

26,274 次观看 • 9 个月前

SAM 3D enables accurate 3D reconstruction from a single image, supporting real-world applications in editing, robotics, and interactive scene generation. Matt, a SAM 3D researcher, explains how the two-model design makes this possible for both people and complex environments. 🔗 Read the SAM 3D Objects research paper: 🔗 Read the SAM 3D Body research paper:

SAM 3D enables accurate 3D reconstruction from a single image, supporting real-world applications in editing, robotics, and interactive scene generation. Matt, a SAM 3D researcher, explains how the two-model design makes this possible for both people and complex environments. 🔗 Read the SAM 3D Objects research paper: 🔗 Read the SAM 3D Body research paper:

AI at Meta

17,858 次观看 • 7 个月前

EgoLifter Open-world 3D Segmentation for Egocentric Perception In this paper we present EgoLifter, a novel system that can automatically segment scenes captured from egocentric sensors into a complete decomposition of individual 3D objects. The system is specifically

EgoLifter Open-world 3D Segmentation for Egocentric Perception In this paper we present EgoLifter, a novel system that can automatically segment scenes captured from egocentric sensors into a complete decomposition of individual 3D objects. The system is specifically

AK

41,117 次观看 • 2 年前

🎤🎤 Excited to introduce COME-robot🤖🤖, Closed-Loop Open-Vocabulary Mobile Manipulation with GPT-4V. It is the first closed-loop framework utilizing the vision-language foundation model for open-ended reasoning and adaptive planning in real-world scenarios. COME-robot demonstrates a significant improvement in task success rate (~25%) compared to SOTA methods. Project: Arxiv:

🎤🎤 Excited to introduce COME-robot🤖🤖, Closed-Loop Open-Vocabulary Mobile Manipulation with GPT-4V. It is the first closed-loop framework utilizing the vision-language foundation model for open-ended reasoning and adaptive planning in real-world scenarios. COME-robot demonstrates a significant improvement in task success rate (~25%) compared to SOTA methods. Project: Arxiv:

Siyuan Huang

22,291 次观看 • 2 年前

We're making a challenging game with many mechanics, numerous possibilities, dangerous enemies, and a fragile protagonist. Additionally, the game features complex platforming sections that require proper planning and the simultaneous use of multiple abilities. #AmongTheWaves #indiegames #gaming

We're making a challenging game with many mechanics, numerous possibilities, dangerous enemies, and a fragile protagonist. Additionally, the game features complex platforming sections that require proper planning and the simultaneous use of multiple abilities. #AmongTheWaves #indiegames #gaming

Among The Waves 💚 Wishlist on Steam!

16,872 次观看 • 3 个月前

🚀 Meta FAIR is releasing several new research artifacts on our road to advanced machine intelligence (AMI). These latest advancements are transforming our understanding of perception. 1️⃣ Meta Perception Encoder: A large-scale vision encoder that excels across several image & video tasks. 2️⃣ Meta Perception Language Model: A fully open & reproducible vision-language model designed to tackle visual recognition tasks. 3️⃣ Meta Locate 3D: An end-to-end model for accurate object localization in 3D environments. 4️⃣ Releasing model weights for our 8B-parameter Dynamic Byte Latent Transformer, an alternative to traditional tokenization methods with the potential to redefine the standards for language model efficiency and reliability. 5️⃣Collaborative Reasoner: A framework for evaluating & improving collaborative reasoning skills in language models. Download the code, datasets, and research papers and learn more about how these artifacts are paving the way for more efficient and accurate AI systems.➡️

🚀 Meta FAIR is releasing several new research artifacts on our road to advanced machine intelligence (AMI). These latest advancements are transforming our understanding of perception. 1️⃣ Meta Perception Encoder: A large-scale vision encoder that excels across several image & video tasks. 2️⃣ Meta Perception Language Model: A fully open & reproducible vision-language model designed to tackle visual recognition tasks. 3️⃣ Meta Locate 3D: An end-to-end model for accurate object localization in 3D environments. 4️⃣ Releasing model weights for our 8B-parameter Dynamic Byte Latent Transformer, an alternative to traditional tokenization methods with the potential to redefine the standards for language model efficiency and reliability. 5️⃣Collaborative Reasoner: A framework for evaluating & improving collaborative reasoning skills in language models. Download the code, datasets, and research papers and learn more about how these artifacts are paving the way for more efficient and accurate AI systems.➡️

AI at Meta

163,214 次观看 • 1 年前

3D-VLA A 3D Vision-Language-Action Generative World Model Recent vision-language-action (VLA) models rely on 2D inputs, lacking integration with the broader realm of the 3D physical world. Furthermore, they perform action prediction by learning a direct mapping from

3D-VLA A 3D Vision-Language-Action Generative World Model Recent vision-language-action (VLA) models rely on 2D inputs, lacking integration with the broader realm of the 3D physical world. Furthermore, they perform action prediction by learning a direct mapping from

AK

84,940 次观看 • 2 年前