Загрузка видео...

Не удалось загрузить видео

Возникла проблема при загрузке этого видео. Это может быть связано с временными проблемами сети или видео может быть недоступно.

На главную

🚀 Introducing InterDyn — our newly accepted CVPR work that explores controllable synthesis of interactive dynamics! Building upon powerful video diffusion models, InterDyn infers future motion and interactions directly from an input image and a dynamic control signal (e.g., a moving hand mask). Check out how we push the... show more

Haven (Haiwen) Feng

1,139 subscribers

44,898 просмотров • 1 год назад •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

Комментарии: 10

Фото профиля Haven (Haiwen) Feng

Haven (Haiwen) Feng1 год назад

Dynamic Control & Beyond: Unlike prior methods that rely on explicit simulation or only static state transitions, InterDyn built a dynamic control branch on top of Stable Video Diffusion. We then fine-tune it to generate complex interactions (e.g. hand-object manipulations) and realistic multi-object collisions without heavy simulation computations. #StableDiffusion #StabilityAI 🧵2/6

Фото профиля Haven (Haiwen) Feng

Haven (Haiwen) Feng1 год назад

Intuitive Physics: At its core, InterDyn showcases the diffusion model’s “knowledge” of real-world physics and causal effects. By simply providing a moving object mask, the system implicitly models collisions, force propagation, and object dynamics—no 3D reconstruction or separate physics engine needed. 🧵3/6

Фото профиля Haven (Haiwen) Feng

Haven (Haiwen) Feng1 год назад

Superior Performance: We evaluate InterDyn on both synthetic (CLEVRER) and real-world datasets (Something-Something-v2), achieving up to 37.5% improvement on LPIPS and 77% on FVD over prior work. Whether it’s multi-object collisions or hand-object manipulations, InterDyn produces diverse and physically plausible videos. 🧵4/6

Фото профиля Haven (Haiwen) Feng

Haven (Haiwen) Feng1 год назад

Toward Interactive Video Generation: This new perspective merges intuitive physics with large-scale generative models, opening the door to controllable dynamics synthesis in complex scenes. We believe InterDyn lays the groundwork for future explorations in interactive video generation. Stay tuned for more! 🧵5/6

Фото профиля Haven (Haiwen) Feng

Haven (Haiwen) Feng1 год назад

This work was co-lead by me and our talented Master intern @rick_akker25502 (He's applying for PhD now, hire him!) together with amazing advisors, @Michael_J_Black , @dimtzionas and @vfabrevaya . More details & demos coming soon! See you in Nashville! #CVPR2025 #AI #ResearchPapers 🧵6/6

Фото профиля Adam

Adam1 год назад

Great work! I had a similar idea for hand-object interaction with video generation but with 3D conditioning

Фото профиля Robert Scoble

Robert Scoble1 год назад

Wow great work!

Фото профиля Kangfu Mei

Kangfu Mei1 год назад

Very nice and creative work!

Фото профиля Daniel Sungho Jung

Daniel Sungho Jung1 год назад

Interesting work! Were there any challenges during the research?

Фото профиля Erika S

Erika S1 год назад

InterDyn’s approach to controllable synthesis of interactive dynamics is fascinating. I’m excited to see how it advances intuitive physics with video generative models—truly pushing boundaries in AI and computer vision!

Похожие видео

📢 Lyra: Generative 3D Scene Reconstruction via Video Diffusion Model Self-Distillation Got only one or a few images and wondering if recovering the 3D environment is a reconstruction or generation problem? Why not do it with a generative reconstruction model! We show that a camera-conditioned video diffusion model can be transformed into a generative reconstruction model that directly outputs a high-quality 3D Gaussian Splatting representation through self-distillation, without requiring real-world training data. Check out our results in the video (wait for dynamic scenes in the second half!) : Project Page: Code and Models: Paper:

📢 Lyra: Generative 3D Scene Reconstruction via Video Diffusion Model Self-Distillation Got only one or a few images and wondering if recovering the 3D environment is a reconstruction or generation problem? Why not do it with a generative reconstruction model! We show that a camera-conditioned video diffusion model can be transformed into a generative reconstruction model that directly outputs a high-quality 3D Gaussian Splatting representation through self-distillation, without requiring real-world training data. Check out our results in the video (wait for dynamic scenes in the second half!) : Project Page: Code and Models: Paper:

Sherwin Bahmani

66,590 просмотров • 9 месяцев назад

Glad that our work “Inference-Time Enhancement of Generative Robot Policies via Predictive World Modeling”, led by Han Qi, has been accepted to IEEE Robotics and Automation Letters! 🎉 We propose Generative Predictive Control (GPC): sample action proposals from a pretrained diffusion policy (“look back”), roll them out with a diffusion-based action-conditioned video world model (“look forward”), then rank or optimize the actions using either a learned reward model or VLM preferences. Conceptually, this is trajectory optimization / MPC with hybrid sampling + gradient optimization, interpreted through modern diffusion priors and video world models. Interestingly, we first posted the paper on arXiv in Feb 2025, when action-conditioned video world models for planning were still rare—now this direction is rapidly gaining traction. Still many open questions, e.g., • how to avoid local minima in planning • what representations work best for world models • how to balance physics priors vs. data-driven learning Paper:

Glad that our work “Inference-Time Enhancement of Generative Robot Policies via Predictive World Modeling”, led by Han Qi, has been accepted to IEEE Robotics and Automation Letters! 🎉 We propose Generative Predictive Control (GPC): sample action proposals from a pretrained diffusion policy (“look back”), roll them out with a diffusion-based action-conditioned video world model (“look forward”), then rank or optimize the actions using either a learned reward model or VLM preferences. Conceptually, this is trajectory optimization / MPC with hybrid sampling + gradient optimization, interpreted through modern diffusion priors and video world models. Interestingly, we first posted the paper on arXiv in Feb 2025, when action-conditioned video world models for planning were still rare—now this direction is rapidly gaining traction. Still many open questions, e.g., • how to avoid local minima in planning • what representations work best for world models • how to balance physics priors vs. data-driven learning Paper:

Heng Yang

18,994 просмотров • 4 месяцев назад

Punchline: World models == VQA (about the future)! Planning with world models can be powerful for robotics/control. But most world models are video generators trained to predict everything, including irrelevant pixels and distractions. We ask - what if a world model only predicted the semantic information necessary for decision-making? Introducing Semantic World Models (SWM). Given an observation and an action sequence, SWMs cast modeling as answering textual questions about the future outcome resulting from the actions. Recasting world modeling as a VQA problem lets us directly leverage the pretrained knowledge and machinery of VLMs for generalizable modeling. We had a lot of fun thinking about how this work helps connect these two seemingly very different fields of study - VLMs and world models! 🧵(1/6) Paper: Fun demo:

Punchline: World models == VQA (about the future)! Planning with world models can be powerful for robotics/control. But most world models are video generators trained to predict everything, including irrelevant pixels and distractions. We ask - what if a world model only predicted the semantic information necessary for decision-making? Introducing Semantic World Models (SWM). Given an observation and an action sequence, SWMs cast modeling as answering textual questions about the future outcome resulting from the actions. Recasting world modeling as a VQA problem lets us directly leverage the pretrained knowledge and machinery of VLMs for generalizable modeling. We had a lot of fun thinking about how this work helps connect these two seemingly very different fields of study - VLMs and world models! 🧵(1/6) Paper: Fun demo:

Abhishek Gupta

61,208 просмотров • 9 месяцев назад

Stable Virtual Camera: Generative View Synthesis with Diffusion Models Hint: Check the project website. It is awesome! Contributions: 1. A training strategy for jointly modeling large viewpoint changes and temporal smoothness. 2. A two-pass procedural sampling method for smooth video generation along arbitrarily long camera trajectories. 3. A comprehensive benchmark that evaluates NVS methods across different datasets and settings. 4. An open-source release of model weights to support future research.

Stable Virtual Camera: Generative View Synthesis with Diffusion Models Hint: Check the project website. It is awesome! Contributions: 1. A training strategy for jointly modeling large viewpoint changes and temporal smoothness. 2. A two-pass procedural sampling method for smooth video generation along arbitrarily long camera trajectories. 3. A comprehensive benchmark that evaluates NVS methods across different datasets and settings. 4. An open-source release of model weights to support future research.

MrNeRF

38,619 просмотров • 1 год назад

Collaborative Score Distillation for Consistent Visual Synthesis paper page: Generative priors of large-scale text-to-image diffusion models enable a wide range of new generation and editing applications on diverse visual modalities. However, when adapting these priors to complex visual modalities, often represented as multiple images (e.g., video), achieving consistency across a set of images is challenging. In this paper, we address this challenge with a novel method, Collaborative Score Distillation (CSD). CSD is based on the Stein Variational Gradient Descent (SVGD). Specifically, we propose to consider multiple samples as "particles" in the SVGD update and combine their score functions to distill generative priors over a set of images synchronously. Thus, CSD facilitates seamless integration of information across 2D images, leading to a consistent visual synthesis across multiple samples. We show the effectiveness of CSD in a variety of tasks, encompassing the visual editing of panorama images, videos, and 3D scenes. Our results underline the competency of CSD as a versatile method for enhancing inter-sample consistency, thereby broadening the applicability of text-to-image diffusion models.

Collaborative Score Distillation for Consistent Visual Synthesis paper page: Generative priors of large-scale text-to-image diffusion models enable a wide range of new generation and editing applications on diverse visual modalities. However, when adapting these priors to complex visual modalities, often represented as multiple images (e.g., video), achieving consistency across a set of images is challenging. In this paper, we address this challenge with a novel method, Collaborative Score Distillation (CSD). CSD is based on the Stein Variational Gradient Descent (SVGD). Specifically, we propose to consider multiple samples as "particles" in the SVGD update and combine their score functions to distill generative priors over a set of images synchronously. Thus, CSD facilitates seamless integration of information across 2D images, leading to a consistent visual synthesis across multiple samples. We show the effectiveness of CSD in a variety of tasks, encompassing the visual editing of panorama images, videos, and 3D scenes. Our results underline the competency of CSD as a versatile method for enhancing inter-sample consistency, thereby broadening the applicability of text-to-image diffusion models.

AK

33,500 просмотров • 3 лет назад

V3D Video Diffusion Models are Effective 3D Generators Automatic 3D generation has recently attracted widespread attention. Recent methods have greatly accelerated the generation speed, but usually produce less-detailed objects due to limited model capacity or 3D data. Motivated by recent advancements in video diffusion models, we introduce V3D, which leverages the world simulation capacity of pre-trained video diffusion models to facilitate 3D generation. To fully unleash the potential of video diffusion to perceive the 3D world, we further introduce geometrical consistency prior and extend the video diffusion model to a multi-view consistent 3D generator. Benefiting from this, the state-of-the-art video diffusion model could be fine-tuned to generate 360degree orbit frames surrounding an object given a single image. With our tailored reconstruction pipelines, we can generate high-quality meshes or 3D Gaussians within 3 minutes. Furthermore, our method can be extended to scene-level novel view synthesis, achieving precise control over the camera path with sparse input views. Extensive experiments demonstrate the superior performance of the proposed approach, especially in terms of generation quality and multi-view consistency

V3D Video Diffusion Models are Effective 3D Generators Automatic 3D generation has recently attracted widespread attention. Recent methods have greatly accelerated the generation speed, but usually produce less-detailed objects due to limited model capacity or 3D data. Motivated by recent advancements in video diffusion models, we introduce V3D, which leverages the world simulation capacity of pre-trained video diffusion models to facilitate 3D generation. To fully unleash the potential of video diffusion to perceive the 3D world, we further introduce geometrical consistency prior and extend the video diffusion model to a multi-view consistent 3D generator. Benefiting from this, the state-of-the-art video diffusion model could be fine-tuned to generate 360degree orbit frames surrounding an object given a single image. With our tailored reconstruction pipelines, we can generate high-quality meshes or 3D Gaussians within 3 minutes. Furthermore, our method can be extended to scene-level novel view synthesis, achieving precise control over the camera path with sparse input views. Extensive experiments demonstrate the superior performance of the proposed approach, especially in terms of generation quality and multi-view consistency

AK

31,997 просмотров • 2 лет назад

We present EgoReAct: Real-time 3D human reaction generation from streaming egocentric video. 🌟Reacting to streaming egocentric video is something humans do every day. We hope EgoReAct makes human motion more human-like. 🔎 What we found: existing ego-reaction data can be spatially inconsistent (e.g., moving reactions paired with fixed-camera videos), which breaks 3D grounding. 📷 What we built: HRD, a spatially aligned egocentric video–reaction dataset (3,500 pairs, 32 categories), plus a spatially aligned ViMo fix for fair evaluation. (Instead of collecting expensive ground-truth motion, we employ VDM to generate the egocentric videos.) 👁️⚡🏃 Our simple yet effective pipeline: motion tokenization for compact discrete codes + an autoregressive Transformer for online, strictly-causal generation. Metric depth and head dynamics further improve 3D spatial consistency. Project Page: ArXiv: #HumanMotion #EgocentricVision #3D #ARVR #Animation #AIGC #DeepLearning #GenerativeAI #Graphics #ComputerVision #Motion

We present EgoReAct: Real-time 3D human reaction generation from streaming egocentric video. 🌟Reacting to streaming egocentric video is something humans do every day. We hope EgoReAct makes human motion more human-like. 🔎 What we found: existing ego-reaction data can be spatially inconsistent (e.g., moving reactions paired with fixed-camera videos), which breaks 3D grounding. 📷 What we built: HRD, a spatially aligned egocentric video–reaction dataset (3,500 pairs, 32 categories), plus a spatially aligned ViMo fix for fair evaluation. (Instead of collecting expensive ground-truth motion, we employ VDM to generate the egocentric videos.) 👁️⚡🏃 Our simple yet effective pipeline: motion tokenization for compact discrete codes + an autoregressive Transformer for online, strictly-causal generation. Metric depth and head dynamics further improve 3D spatial consistency. Project Page: ArXiv: #HumanMotion #EgocentricVision #3D #ARVR #Animation #AIGC #DeepLearning #GenerativeAI #Graphics #ComputerVision #Motion

Zhiyang (Frank) Dou

11,336 просмотров • 6 месяцев назад

Is VideoGen starting to become good enough for robotic manipulation? 🤖 Check out our recent work, RIGVid — Robots Imitating Generated Videos — where we use AI-generated videos as intermediate representations and 6-DoF motion retargeting to guide robots in diverse manipulation tasks: pouring, wiping, mixing, and more. 🔗 Key takeaways: - VideoGen starts to become good enough for robotics - As the field progresses, we are expecting much better results in the coming years - Depending on whether video prediction models take actions or not (VideoGen vs Action-Conditioned Video Prediction), there are different ways to use them. - Controllability & steerability are still issues In the paper, we explore: – How do different video generation models compare for robotic imitation? – Can generated videos replace real videos for imitation? – What causes failure of imitation given high-quality videos? – How does imitating from video compare with other representations (e.g., keypoint constraints like ReKep)? 🎥 Watch the video for (1) AI-generated inputs, (2) robot executions, and (3) the 3D intermediate representation bridging the embodiment gap.

Is VideoGen starting to become good enough for robotic manipulation? 🤖 Check out our recent work, RIGVid — Robots Imitating Generated Videos — where we use AI-generated videos as intermediate representations and 6-DoF motion retargeting to guide robots in diverse manipulation tasks: pouring, wiping, mixing, and more. 🔗 Key takeaways: - VideoGen starts to become good enough for robotics - As the field progresses, we are expecting much better results in the coming years - Depending on whether video prediction models take actions or not (VideoGen vs Action-Conditioned Video Prediction), there are different ways to use them. - Controllability & steerability are still issues In the paper, we explore: – How do different video generation models compare for robotic imitation? – Can generated videos replace real videos for imitation? – What causes failure of imitation given high-quality videos? – How does imitating from video compare with other representations (e.g., keypoint constraints like ReKep)? 🎥 Watch the video for (1) AI-generated inputs, (2) robot executions, and (3) the 3D intermediate representation bridging the embodiment gap.

Yunzhu Li

16,540 просмотров • 1 год назад

Multi-Track Timeline Control for Text-Driven 3D Human Motion Generation paper page: Recent advances in generative modeling have led to promising progress on synthesizing 3D human motion from text, with methods that can generate character animations from short prompts and specified durations. However, using a single text prompt as input lacks the fine-grained control needed by animators, such as composing multiple actions and defining precise durations for parts of the motion. To address this, we introduce the new problem of timeline control for text-driven motion synthesis, which provides an intuitive, yet fine-grained, input interface for users. Instead of a single prompt, users can specify a multi-track timeline of multiple prompts organized in temporal intervals that may overlap. This enables specifying the exact timings of each action and composing multiple actions in sequence or at overlapping intervals. To generate composite animations from a multi-track timeline, we propose a new test-time denoising method. This method can be integrated with any pre-trained motion diffusion model to synthesize realistic motions that accurately reflect the timeline. At every step of denoising, our method processes each timeline interval (text prompt) individually, subsequently aggregating the predictions with consideration for the specific body parts engaged in each action. Experimental comparisons and ablations validate that our method produces realistic motions that respect the semantics and timing of given text prompts.

Multi-Track Timeline Control for Text-Driven 3D Human Motion Generation paper page: Recent advances in generative modeling have led to promising progress on synthesizing 3D human motion from text, with methods that can generate character animations from short prompts and specified durations. However, using a single text prompt as input lacks the fine-grained control needed by animators, such as composing multiple actions and defining precise durations for parts of the motion. To address this, we introduce the new problem of timeline control for text-driven motion synthesis, which provides an intuitive, yet fine-grained, input interface for users. Instead of a single prompt, users can specify a multi-track timeline of multiple prompts organized in temporal intervals that may overlap. This enables specifying the exact timings of each action and composing multiple actions in sequence or at overlapping intervals. To generate composite animations from a multi-track timeline, we propose a new test-time denoising method. This method can be integrated with any pre-trained motion diffusion model to synthesize realistic motions that accurately reflect the timeline. At every step of denoising, our method processes each timeline interval (text prompt) individually, subsequently aggregating the predictions with consideration for the specific body parts engaged in each action. Experimental comparisons and ablations validate that our method produces realistic motions that respect the semantics and timing of given text prompts.

AK

126,585 просмотров • 2 лет назад

InstantDrag Improving Interactivity in Drag-based Image Editing discuss: Drag-based image editing has recently gained popularity for its interactivity and precision. However, despite the ability of text-to-image models to generate samples within a second, drag editing still lags behind due to the challenge of accurately reflecting user interaction while maintaining image content. Some existing approaches rely on computationally intensive per-image optimization or intricate guidance-based methods, requiring additional inputs such as masks for movable regions and text prompts, thereby compromising the interactivity of the editing process. We introduce InstantDrag, an optimization-free pipeline that enhances interactivity and speed, requiring only an image and a drag instruction as input. InstantDrag consists of two carefully designed networks: a drag-conditioned optical flow generator (FlowGen) and an optical flow-conditioned diffusion model (FlowDiffusion). InstantDrag learns motion dynamics for drag-based image editing in real-world video datasets by decomposing the task into motion generation and motion-conditioned image generation. We demonstrate InstantDrag's capability to perform fast, photo-realistic edits without masks or text prompts through experiments on facial video datasets and general scenes. These results highlight the efficiency of our approach in handling drag-based image editing, making it a promising solution for interactive, real-time applications.

InstantDrag Improving Interactivity in Drag-based Image Editing discuss: Drag-based image editing has recently gained popularity for its interactivity and precision. However, despite the ability of text-to-image models to generate samples within a second, drag editing still lags behind due to the challenge of accurately reflecting user interaction while maintaining image content. Some existing approaches rely on computationally intensive per-image optimization or intricate guidance-based methods, requiring additional inputs such as masks for movable regions and text prompts, thereby compromising the interactivity of the editing process. We introduce InstantDrag, an optimization-free pipeline that enhances interactivity and speed, requiring only an image and a drag instruction as input. InstantDrag consists of two carefully designed networks: a drag-conditioned optical flow generator (FlowGen) and an optical flow-conditioned diffusion model (FlowDiffusion). InstantDrag learns motion dynamics for drag-based image editing in real-world video datasets by decomposing the task into motion generation and motion-conditioned image generation. We demonstrate InstantDrag's capability to perform fast, photo-realistic edits without masks or text prompts through experiments on facial video datasets and general scenes. These results highlight the efficiency of our approach in handling drag-based image editing, making it a promising solution for interactive, real-time applications.

AK

71,232 просмотров • 1 год назад

🇨🇳 Another great Chinese Model, OmniHuman-1.5 from ByteDance Turns 1 image plus a voice track into expressive avatar video by pairing a System 1 and System 2 inspired planner with a Diffusion Transformer, Produces coherent motion for over 1 minute with moving camera and multi character scenes. Most avatar models move to the beat of the audio but miss meaning, so gestures feel generic and emotions feel shallow. The fix here is a Multimodal LLM planner that listens to the speech and drafts a structured plan describing intent, emotions, beats, and high level actions, which gives the motion engine clear semantic targets instead of only rhythm. The motion engine is a Multimodal Diffusion Transformer that fuses the plan with audio, the single reference image, and optional text prompts, then synthesizes continuous body, face, and head motion that matches both words and tone. A key trick is a Pseudo Last Frame, a synthetic target that summarizes the next expected state, which stabilizes fusion across modalities and keeps motion consistent over long spans. From just 1 image and speech, the system outputs speaking avatars with synchronized lips, context aware gestures, and continuous camera movement, and it also supports multi character interactions without manual choreography. Reported results show strong lip sync accuracy, high video quality, natural motion, and close match to text prompts, and the same setup works on nonhuman characters too.

Rohan Paul

63,859 просмотров • 10 месяцев назад

Alibaba presents MIMO Controllable Character Video Synthesis with Spatial Decomposed Modeling Character video synthesis aims to produce realistic videos of animatable characters within lifelike scenes. As a fundamental problem in the computer vision and graphics community, 3D works typically require multi-view captures for per-case training, which severely limits their applicability of modeling arbitrary characters in a short time. Recent 2D methods break this limitation via pre-trained diffusion models, but they struggle for pose generality and scene interaction. To this end, we propose MIMO, a novel framework which can not only synthesize character videos with controllable attributes (i.e., character, motion and scene) provided by simple user inputs, but also simultaneously achieve advanced scalability to arbitrary characters, generality to novel 3D motions, and applicability to interactive real-world scenes in a unified framework. The core idea is to encode the 2D video to compact spatial codes, considering the inherent 3D nature of video occurrence. Concretely, we lift the 2D frame pixels into 3D using monocular depth estimators, and decompose the video clip to three spatial components (i.e., main human, underlying scene, and floating occlusion) in hierarchical layers based on the 3D depth. These components are further encoded to canonical identity code, structured motion code and full scene code, which are utilized as control signals of synthesis process. The design of spatial decomposed modeling enables flexible user control, complex motion expression, as well as 3D-aware synthesis for scene interactions. Experimental results demonstrate effectiveness and robustness of the proposed method.

Alibaba presents MIMO Controllable Character Video Synthesis with Spatial Decomposed Modeling Character video synthesis aims to produce realistic videos of animatable characters within lifelike scenes. As a fundamental problem in the computer vision and graphics community, 3D works typically require multi-view captures for per-case training, which severely limits their applicability of modeling arbitrary characters in a short time. Recent 2D methods break this limitation via pre-trained diffusion models, but they struggle for pose generality and scene interaction. To this end, we propose MIMO, a novel framework which can not only synthesize character videos with controllable attributes (i.e., character, motion and scene) provided by simple user inputs, but also simultaneously achieve advanced scalability to arbitrary characters, generality to novel 3D motions, and applicability to interactive real-world scenes in a unified framework. The core idea is to encode the 2D video to compact spatial codes, considering the inherent 3D nature of video occurrence. Concretely, we lift the 2D frame pixels into 3D using monocular depth estimators, and decompose the video clip to three spatial components (i.e., main human, underlying scene, and floating occlusion) in hierarchical layers based on the 3D depth. These components are further encoded to canonical identity code, structured motion code and full scene code, which are utilized as control signals of synthesis process. The design of spatial decomposed modeling enables flexible user control, complex motion expression, as well as 3D-aware synthesis for scene interactions. Experimental results demonstrate effectiveness and robustness of the proposed method.

AK

148,998 просмотров • 1 год назад

🚀We are thrilled to open-source Hunyuan-GameCraft, a high-dynamic interactive game video generation framework built on HunyuanVideo. It generates playable and physically realistic videos from a single scene image and user action signals, empowering creators and developers to "direct" games with first-person or third-person perspectives. Key Advantages: 🔹High Dynamics: Unifies standard keyboard inputs into a shared continuous action space, enabling high-precision control over velocity and angle. This allows for the exploration of complex trajectories, overcoming the stiff, limited motion of traditional models. It can also generate dynamic environmental content like moving clouds, rain, snow, and water flow. 🔹Long-term Consistency: Uses hybrid history condition to preserve the original scene information after significant movement. 🔹Significant Cost Reduction: No need for expensive modeling/rendering. PCM distillation compresses inference steps, boosting speed and lowering costs. This allows the quantized 13B model to run on consumer-grade GPUs like the RTX 4090. Project Page: Code: Technical Report: Hugging Face：

🚀We are thrilled to open-source Hunyuan-GameCraft, a high-dynamic interactive game video generation framework built on HunyuanVideo. It generates playable and physically realistic videos from a single scene image and user action signals, empowering creators and developers to "direct" games with first-person or third-person perspectives. Key Advantages: 🔹High Dynamics: Unifies standard keyboard inputs into a shared continuous action space, enabling high-precision control over velocity and angle. This allows for the exploration of complex trajectories, overcoming the stiff, limited motion of traditional models. It can also generate dynamic environmental content like moving clouds, rain, snow, and water flow. 🔹Long-term Consistency: Uses hybrid history condition to preserve the original scene information after significant movement. 🔹Significant Cost Reduction: No need for expensive modeling/rendering. PCM distillation compresses inference steps, boosting speed and lowering costs. This allows the quantized 13B model to run on consumer-grade GPUs like the RTX 4090. Project Page: Code: Technical Report: Hugging Face：

Tencent Hy

169,638 просмотров • 11 месяцев назад

Robots might learn better from video than from language! 📼 Most Vision-Language-Action (VLA) models learn what to do from text, but still struggle with how things move in the real world. That makes them data-hungry and slow to train. mimic video takes a different route. Instead of grounding robot control in text, it grounds it in video, using large pre-trained video models that already capture physical motion and dynamics. The idea is straightforward: let the video model handle “what will happen next,” and let a smaller control model focus only on turning that visual plan into robot actions. The result is big gains in practice. Robots trained this way need 10× less data, converge twice as fast, and perform better on both simulated benchmarks and real bimanual manipulation tasks. If robots can “imagine” motion using video, control becomes a much simpler problem. Shoutout to Jonas Pai, Liam Achenbach, Oier Mees, Elvis Nava and the rest of the team! Here's the project page: ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Robots might learn better from video than from language! 📼 Most Vision-Language-Action (VLA) models learn what to do from text, but still struggle with how things move in the real world. That makes them data-hungry and slow to train. mimic video takes a different route. Instead of grounding robot control in text, it grounds it in video, using large pre-trained video models that already capture physical motion and dynamics. The idea is straightforward: let the video model handle “what will happen next,” and let a smaller control model focus only on turning that visual plan into robot actions. The result is big gains in practice. Robots trained this way need 10× less data, converge twice as fast, and perform better on both simulated benchmarks and real bimanual manipulation tasks. If robots can “imagine” motion using video, control becomes a much simpler problem. Shoutout to Jonas Pai, Liam Achenbach, Oier Mees, Elvis Nava and the rest of the team! Here's the project page: ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Lukas Ziegler

49,920 просмотров • 7 месяцев назад

📢 Our lab has been exploring 3D world models for years — and we’re thrilled to share **PhysTwin**: a milestone that reconstructs object appearance, geometry, and dynamics from just a few seconds of interaction! Led by the amazing Hanxiao Jiang 👉 PhysTwin combines **Gaussian splatting** with **inverse dynamics optimization** based on simple **spring-mass** systems. ⚙️ The result? Real-time, action-conditioned 3D video prediction under novel interactions (i.e., 3D world models). 🔑 A few key takeaways: 1. Having the right structure (e.g., particles/masses) helps navigate the trade-off between sample efficiency, generalization, and broad applicability. 2. Visual foundation models (VFMs) have matured to the point where they can provide rich supervision for world modeling (e.g., tracking, shape completion). 3. Beyond VFMs, many crucial components have come together in recent years: Gaussian splats for rendering, NVIDIA Warp for high-performance simulation, and scene/asset generation from a wide range of labs and companies. The future of 3D world models is looking bright! ✨ 4. The resulting digital twin supports a wide range of downstream applications—especially in data generation and policy evaluation, thanks to its realistic rendering and simulation capabilities. 🎥 All code and data to reproduce the results, along with interactive demos, are available on the website. Check the following visualizations of: (1) observations, (2) reconstructed state/actions, (3) interactive digital twins, and (4) the overlays between real-world robot teleoperation and our model’s open-loop predictions.

📢 Our lab has been exploring 3D world models for years — and we’re thrilled to share PhysTwin: a milestone that reconstructs object appearance, geometry, and dynamics from just a few seconds of interaction! Led by the amazing Hanxiao Jiang 👉 PhysTwin combines Gaussian splatting with inverse dynamics optimization based on simple spring-mass systems. ⚙️ The result? Real-time, action-conditioned 3D video prediction under novel interactions (i.e., 3D world models). 🔑 A few key takeaways: 1. Having the right structure (e.g., particles/masses) helps navigate the trade-off between sample efficiency, generalization, and broad applicability. 2. Visual foundation models (VFMs) have matured to the point where they can provide rich supervision for world modeling (e.g., tracking, shape completion). 3. Beyond VFMs, many crucial components have come together in recent years: Gaussian splats for rendering, NVIDIA Warp for high-performance simulation, and scene/asset generation from a wide range of labs and companies. The future of 3D world models is looking bright! ✨ 4. The resulting digital twin supports a wide range of downstream applications—especially in data generation and policy evaluation, thanks to its realistic rendering and simulation capabilities. 🎥 All code and data to reproduce the results, along with interactive demos, are available on the website. Check the following visualizations of: (1) observations, (2) reconstructed state/actions, (3) interactive digital twins, and (4) the overlays between real-world robot teleoperation and our model’s open-loop predictions.

Yunzhu Li

25,279 просмотров • 1 год назад

We benchmarked leading multimodal foundation models (GPT-4o, Claude 3.5 Sonnet, Gemini, Llama, etc.) on standard computer vision tasks—from segmentation to surface normal estimation—using standard datasets like COCO and ImageNet. These models have made remarkable progress; however, it is unclear exactly where they stand in terms of understanding vision in detail. Especially when it comes to tasks beyond question-answering. How well do they understand an object's segments or geometry? Our analyses yield an assessment that is quantitatively and qualitatively detailed and is compatible with evaluations developed in the field of computer vision over the past decades. Observed trends: 🔹 The foundation models consistently underperform task-specific SOTA models across all tasks. However, they are respectable generalists, which is remarkable as they are presumably trained primarily on image-text-based tasks. 🔹 They perform semantic tasks notably better than geometric ones. 🔹 GPT-4o performs the best among non-reasoning models, getting the top position in 4 out of 6 tasks. 🔹 Reasoning models, e.g., o3, show improvements in geometric tasks. 🔹 The 'image generation' models, e.g., GPT-40 Image Generation, which have been natively trained multimodally, exhibit quirks. E.g., hallucinated objects, misalignment between the input and output, etc. 🔹 While the prompting techniques affect performance, better models exhibit less sensitivity to variations in prompts. We control for the variance introduced by the prompting methods in our experiments. 🌐 Detailed analyses, visualizations: ⌨️ code: 🧵 1/n

We benchmarked leading multimodal foundation models (GPT-4o, Claude 3.5 Sonnet, Gemini, Llama, etc.) on standard computer vision tasks—from segmentation to surface normal estimation—using standard datasets like COCO and ImageNet. These models have made remarkable progress; however, it is unclear exactly where they stand in terms of understanding vision in detail. Especially when it comes to tasks beyond question-answering. How well do they understand an object's segments or geometry? Our analyses yield an assessment that is quantitatively and qualitatively detailed and is compatible with evaluations developed in the field of computer vision over the past decades. Observed trends: 🔹 The foundation models consistently underperform task-specific SOTA models across all tasks. However, they are respectable generalists, which is remarkable as they are presumably trained primarily on image-text-based tasks. 🔹 They perform semantic tasks notably better than geometric ones. 🔹 GPT-4o performs the best among non-reasoning models, getting the top position in 4 out of 6 tasks. 🔹 Reasoning models, e.g., o3, show improvements in geometric tasks. 🔹 The 'image generation' models, e.g., GPT-40 Image Generation, which have been natively trained multimodally, exhibit quirks. E.g., hallucinated objects, misalignment between the input and output, etc. 🔹 While the prompting techniques affect performance, better models exhibit less sensitivity to variations in prompts. We control for the variance introduced by the prompting methods in our experiments. 🌐 Detailed analyses, visualizations: ⌨️ code: 🧵 1/n

Amir Zamir

73,074 просмотров • 1 год назад

I’ve dreamt of creating a tool that could animate anyone with any motion from just ONE image… and now it’s a reality! 🎉 Super excited to introduce updated 3DHM: Synthesizing Moving People with 3D Control. 🕺💃3DHM can generate human videos from a single real or synthetic human image. #Animation #GenAI #AI #3DHM ✨ The magic of 3D control? Turning 2D pixels into lifelike, animated humans. 🎥 Check out our demo (and Merry Christmas)! Paper: Github: Webpage: Proudly working with the great Junming (Leo) Chen, , Yossi Gandelsman, Alyosha Efros and Jitendra MALIK😃 Kindly note: This video is intended solely for research purposes and is not authorized for commercial use.

I’ve dreamt of creating a tool that could animate anyone with any motion from just ONE image… and now it’s a reality! 🎉 Super excited to introduce updated 3DHM: Synthesizing Moving People with 3D Control. 🕺💃3DHM can generate human videos from a single real or synthetic human image. #Animation #GenAI #AI #3DHM ✨ The magic of 3D control? Turning 2D pixels into lifelike, animated humans. 🎥 Check out our demo (and Merry Christmas)! Paper: Github: Webpage: Proudly working with the great Junming (Leo) Chen, , Yossi Gandelsman, Alyosha Efros and Jitendra MALIK😃 Kindly note: This video is intended solely for research purposes and is not authorized for commercial use.

Boyi Li

52,482 просмотров • 1 год назад

Fast View Synthesis of Casual Videos paper page: Novel view synthesis from an in-the-wild video is difficult due to challenges like scene dynamics and lack of parallax. While existing methods have shown promising results with implicit neural radiance fields, they are slow to train and render. This paper revisits explicit video representations to synthesize high-quality novel views from a monocular video efficiently. We treat static and dynamic video content separately. Specifically, we build a global static scene model using an extended plane-based scene representation to synthesize temporally coherent novel video. Our plane-based scene representation is augmented with spherical harmonics and displacement maps to capture view-dependent effects and model non-planar complex surface geometry. We opt to represent the dynamic content as per-frame point clouds for efficiency. While such representations are inconsistency-prone, minor temporal inconsistencies are perceptually masked due to motion. We develop a method to quickly estimate such a hybrid video representation and render novel views in real time. Our experiments show that our method can render high-quality novel views from an in-the-wild video with comparable quality to state-of-the-art methods while being 100x faster in training and enabling real-time rendering.

Fast View Synthesis of Casual Videos paper page: Novel view synthesis from an in-the-wild video is difficult due to challenges like scene dynamics and lack of parallax. While existing methods have shown promising results with implicit neural radiance fields, they are slow to train and render. This paper revisits explicit video representations to synthesize high-quality novel views from a monocular video efficiently. We treat static and dynamic video content separately. Specifically, we build a global static scene model using an extended plane-based scene representation to synthesize temporally coherent novel video. Our plane-based scene representation is augmented with spherical harmonics and displacement maps to capture view-dependent effects and model non-planar complex surface geometry. We opt to represent the dynamic content as per-frame point clouds for efficiency. While such representations are inconsistency-prone, minor temporal inconsistencies are perceptually masked due to motion. We develop a method to quickly estimate such a hybrid video representation and render novel views in real time. Our experiments show that our method can render high-quality novel views from an in-the-wild video with comparable quality to state-of-the-art methods while being 100x faster in training and enabling real-time rendering.

AK

20,668 просмотров • 2 лет назад

I had to do this of course - test out the new Google Omni Flash model with my favorite 3d art sandbox - Dreams. Just a quick test for now.. I'll be sharing more examples and techniques soon but wanted to show you this one. There's something very fitting about combining "handmade" 3d models and animations sculpted from scratch in software that runs on 2013 hardware, with some of the most advanced AI technology in the world. As part of my work for Google, I've used Google V2V models on a few different projects now - Ancestra, Dear Upstairs Neighbors and most recently one of the IO films (I'll share more about that project soon) 🥰 - and video continues to be my favorite form of input for Generative Ai workflows. If an image says more than a thousand words, a video says more than a million. And knowing that these capabilities are now in the hands of all of you, makes me very excited to see the new workflows and projects that will come out of it. What do you think of Omni so far?

I had to do this of course - test out the new Google Omni Flash model with my favorite 3d art sandbox - Dreams. Just a quick test for now.. I'll be sharing more examples and techniques soon but wanted to show you this one. There's something very fitting about combining "handmade" 3d models and animations sculpted from scratch in software that runs on 2013 hardware, with some of the most advanced AI technology in the world. As part of my work for Google, I've used Google V2V models on a few different projects now - Ancestra, Dear Upstairs Neighbors and most recently one of the IO films (I'll share more about that project soon) 🥰 - and video continues to be my favorite form of input for Generative Ai workflows. If an image says more than a thousand words, a video says more than a million. And knowing that these capabilities are now in the hands of all of you, makes me very excited to see the new workflows and projects that will come out of it. What do you think of Omni so far?

Martin Nebelong

38,323 просмотров • 2 месяцев назад

We’re on the verge of interactive, real-time, photorealistic video generation with what are called World Models. These require a fair amount of expensive compute, but costs will come down over time. Today, these interactions are essentially real-time dreams. They lack the persistence and shared logic that turns a video into a multiplayer game. Or concert. Or classroom. Or the holodeck. We believe the ultimate architecture for gaming (and the holodeck) is Roblox Reality. It’s a hybrid architecture that marries the structured data and logic of the Roblox Engine and Roblox Cloud with the generative power of Video World Models. The Roblox Engine provides the underlying synchronized ground truth—the score, physics, multiplayer sync, etc.—while our video model acts as a Super Upsampler to layer on photorealistic detail. We believe this will ultimately remove barriers to high-fidelity creation, allowing a team of three people to build a narrative-driven, photorealistic masterpiece in a single week. This is an early look at turning solitary AI dreams into a social, playable reality.

We’re on the verge of interactive, real-time, photorealistic video generation with what are called World Models. These require a fair amount of expensive compute, but costs will come down over time. Today, these interactions are essentially real-time dreams. They lack the persistence and shared logic that turns a video into a multiplayer game. Or concert. Or classroom. Or the holodeck. We believe the ultimate architecture for gaming (and the holodeck) is Roblox Reality. It’s a hybrid architecture that marries the structured data and logic of the Roblox Engine and Roblox Cloud with the generative power of Video World Models. The Roblox Engine provides the underlying synchronized ground truth—the score, physics, multiplayer sync, etc.—while our video model acts as a Super Upsampler to layer on photorealistic detail. We believe this will ultimately remove barriers to high-fidelity creation, allowing a team of three people to build a narrative-driven, photorealistic masterpiece in a single week. This is an early look at turning solitary AI dreams into a social, playable reality.

David Baszucki

776,581 просмотров • 2 месяцев назад