Video yükleniyor...

Video Yüklenemedi

Bu video yüklenirken bir sorun oluştu. Bu geçici bir ağ sorunundan kaynaklanıyor olabilir veya video kullanılamıyor olabilir.

Ana Sayfaya Dön

Human3R basically says "what if we stopped treating human reconstruction & scene reconstruction as separate problems." Feed it a casual 2D videos and it spits out multiple people's full body models + 3D scene + camera trajectory - all at once, no multi-stage pipeline required.

Bilawal Sidhu

107,920 subscribers

21,647 görüntüleme • 10 ay önce •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Yorum

Yorum bulunmuyor

Orijinal gönderinin yorumları burada görünecek

Benzer Videolar

SAM 3D Body is a CVPR 2026 award candidate paper from AI at Meta model recovers a full 3D human body mesh from a single RGB image you can run it automatically, or guide the reconstruction with masks and 2D keypoints thx to Niels Rogge for awesome demo idea

SAM 3D Body is a CVPR 2026 award candidate paper from AI at Meta model recovers a full 3D human body mesh from a single RGB image you can run it automatically, or guide the reconstruction with masks and 2D keypoints thx to Niels Rogge for awesome demo idea

SkalskiP

72,500 görüntüleme • 2 ay önce

Human3R: Everyone Everywhere All at Once Note: I recorded the video from the interactive demo on their project page (linked in the comment below). Abstract (excerpt): Human3R jointly recovers global multi-person SMPL-X bodies ("everyone"), dense 3D scenes ("everywhere"), and camera trajectories in a single forward pass ("all-at-once"). Our method builds upon the 4D online reconstruction model CUT3R and uses parameter-efficient visual prompt tuning to preserve CUT3R's rich spatiotemporal priors while enabling direct readout of multiple SMPL-X bodies. Human3R is a unified model that eliminates heavy dependencies and iterative refinement. After being trained on the relatively small-scale synthetic dataset BEDLAM for just one day on one GPU, it achieves superior performance with remarkable efficiency: it reconstructs multiple humans in a one-shot manner, along with 3D scenes, in one stage, at real-time speed (15 FPS) with a low memory footprint (8 GB).

Human3R: Everyone Everywhere All at Once Note: I recorded the video from the interactive demo on their project page (linked in the comment below). Abstract (excerpt): Human3R jointly recovers global multi-person SMPL-X bodies ("everyone"), dense 3D scenes ("everywhere"), and camera trajectories in a single forward pass ("all-at-once"). Our method builds upon the 4D online reconstruction model CUT3R and uses parameter-efficient visual prompt tuning to preserve CUT3R's rich spatiotemporal priors while enabling direct readout of multiple SMPL-X bodies. Human3R is a unified model that eliminates heavy dependencies and iterative refinement. After being trained on the relatively small-scale synthetic dataset BEDLAM for just one day on one GPU, it achieves superior performance with remarkable efficiency: it reconstructs multiple humans in a one-shot manner, along with 3D scenes, in one stage, at real-time speed (15 FPS) with a low memory footprint (8 GB).

MrNeRF

35,783 görüntüleme • 10 ay önce

ReViV reconstructs viewer-centric human motion (body, hand, and gaze) and view-centric scene geometry (camera and depth) from a single egocentric RGB video in a unified feed-forward model. It formulates the task as learning the full joint probability distribution over multimodal signals, including RGB video, camera trajectory, gaze direction, full-body motion, hand motion, and depth. Powered by a Masked Generative Egocentric Transformer, ReViV operates within a single feed-forward architecture to simultaneously reconstruct the temporally consistent 4D reconstruction across the viewer and the view with fast inference speed. Paper Title: ReViV: Reconstructing the Viewer and the View in 4D from Monocular Project: Link:

ReViV reconstructs viewer-centric human motion (body, hand, and gaze) and view-centric scene geometry (camera and depth) from a single egocentric RGB video in a unified feed-forward model. It formulates the task as learning the full joint probability distribution over multimodal signals, including RGB video, camera trajectory, gaze direction, full-body motion, hand motion, and depth. Powered by a Masked Generative Egocentric Transformer, ReViV operates within a single feed-forward architecture to simultaneously reconstruct the temporally consistent 4D reconstruction across the viewer and the view with fast inference speed. Paper Title: ReViV: Reconstructing the Viewer and the View in 4D from Monocular Project: Link:

AI Bites | YouTube Channel

18,590 görüntüleme • 15 gün önce

Wonderland: Navigating 3D Scenes from a Single Image Contributions: • First, we introduce a representation for controllable 3D generation by leveraging the generative priors from camera-guided video diffusion models. Unlike image models, video diffusion models are trained on extensive video datasets. This enables them to capture comprehensive spatial relationships within scenes across multiple views and embed a form of "3D awareness" in their latent space, which allows us to maintain 3D consistency in novel view synthesis. • Second, to achieve controllable novel view generation, we empower video models with precise control over specified camera motions. We introduce a novel dual-branch conditioning mechanism that effectively incorporates desired diverse camera trajectories into the video diffusion model. This enables expansion of a single image into a multi-view consistent capture of a 3D scene with precise pose control. • Third, to achieve efficient 3D reconstruction, we directly transform video latents into 3DGS. We propose a novel latent-based large reconstruction model (LaLRM) that lifts video latents to 3D in a feed-forward manner. With this design, during inference, our model directly predicts 3DGS from a single input image, effectively aligning the generation and reconstruction tasks—and bridging image space and 3D space—through the video latent space. Compared with reconstructing scenes from images, the video latent space offers a 256× spatial-temporal reduction while retaining essential and consistent 3D structural details. Such a high degree of compression is crucial, as it allows the LaLRM to handle a wider range of 3D scenes within the reconstruction framework, with the same memory constraints.

Wonderland: Navigating 3D Scenes from a Single Image Contributions: • First, we introduce a representation for controllable 3D generation by leveraging the generative priors from camera-guided video diffusion models. Unlike image models, video diffusion models are trained on extensive video datasets. This enables them to capture comprehensive spatial relationships within scenes across multiple views and embed a form of "3D awareness" in their latent space, which allows us to maintain 3D consistency in novel view synthesis. • Second, to achieve controllable novel view generation, we empower video models with precise control over specified camera motions. We introduce a novel dual-branch conditioning mechanism that effectively incorporates desired diverse camera trajectories into the video diffusion model. This enables expansion of a single image into a multi-view consistent capture of a 3D scene with precise pose control. • Third, to achieve efficient 3D reconstruction, we directly transform video latents into 3DGS. We propose a novel latent-based large reconstruction model (LaLRM) that lifts video latents to 3D in a feed-forward manner. With this design, during inference, our model directly predicts 3DGS from a single input image, effectively aligning the generation and reconstruction tasks—and bridging image space and 3D space—through the video latent space. Compared with reconstructing scenes from images, the video latent space offers a 256× spatial-temporal reduction while retaining essential and consistent 3D structural details. Such a high degree of compression is crucial, as it allows the LaLRM to handle a wider range of 3D scenes within the reconstruction framework, with the same memory constraints.

MrNeRF

52,849 görüntüleme • 1 yıl önce

I typed one prompt and Dreamina AI turned it into a full cinematic cat fight scene. No VFX. No stock footage. Just Seedream 5.0 Lite and Seedance 2.0 doing the heavy lifting. Slow motion. Rotating camera. Multi-shot transitions. All AI generated. Let me show you what these models can actually do 👇

I typed one prompt and Dreamina AI turned it into a full cinematic cat fight scene. No VFX. No stock footage. Just Seedream 5.0 Lite and Seedance 2.0 doing the heavy lifting. Slow motion. Rotating camera. Multi-shot transitions. All AI generated. Let me show you what these models can actually do 👇

Jibril

184,973 görüntüleme • 4 ay önce

I've been working a lot with SAM3 and the Momentum Human Rig (MHR). I finally integrated it into the data I'm working with Rerun. The progression I've taken looks as follows SAM3 + SAM3D-body on 1. a single image 2. a set of multiple images 3. a single video 4. A multiview video capture I took inspiration from the SAM3D-body paper and built a multiview fitting optimization pipeline. This pipeline involves using the 2D keypoints from the single-view pipeline, triangulating them, and employing an L1 loss between the 2D/3D keypoints. The temporal stability isn't great, so that's the next portion I'm going to focus on. One really frustrating thing about SAM3D-body is the lack of per-joint confidence values. It makes it harder to deal with occlusions. I'm probably going to need to use a separate model, or maybe add a confidence head.

I've been working a lot with SAM3 and the Momentum Human Rig (MHR). I finally integrated it into the data I'm working with Rerun. The progression I've taken looks as follows SAM3 + SAM3D-body on 1. a single image 2. a set of multiple images 3. a single video 4. A multiview video capture I took inspiration from the SAM3D-body paper and built a multiview fitting optimization pipeline. This pipeline involves using the 2D keypoints from the single-view pipeline, triangulating them, and employing an L1 loss between the 2D/3D keypoints. The temporal stability isn't great, so that's the next portion I'm going to focus on. One really frustrating thing about SAM3D-body is the lack of per-joint confidence values. It makes it harder to deal with occlusions. I'm probably going to need to use a separate model, or maybe add a confidence head.

Pablo Vela

42,267 görüntüleme • 6 ay önce

Two weeks ago I fixed one of my teeth with algorithms I wrote a couple of years ago! I got hooked by 3D scanning when I started to work for a software shop in Zurich that was programming 3D computational geometry algorithms for denture scanning to produce crowns (and more). Back then, a typical reconstruction pipeline was like: scan the patient’s teeth using an intraoral scanner, reconstruct the surface mesh, design the restoration digitally, and finally mill the crown out of ceramic. We were working mostly with point clouds and meshes, but it wasn’t just math, it was craftsmanship translated into a digital process. Every micron mattered. You could literally see how a good algorithm meant a better fit in someone’s mouth. Gaussian Splatting isn’t about surface reconstruction, it’s about appearance reconstruction. It doesn’t care about explicit topology, it captures how light interacts with the scene. In a sense, it’s the opposite philosophy of the dental world: instead of modeling what the object is, it models how the object looks. 3D Gaussian Splatting enables applications like training self driving cars, teaching robots to understand their environment, creating virtual worlds, or monitoring real sites. It represents scenes as millions of small Gaussians rendered in real time without the need for meshes or textures. Coming from a world where precision geometry was everything, this shift felt natural. It’s still about reconstruction, but with a different goal: not manufacturing a perfect object, but reproducing how the world actually looks. Two weeks ago I got my first dental crown, made with the same software, reconstruction algorithms, and Swiss precision I once helped develop. I haven’t worked there in two years, but sitting in that chair and seeing the process from the other side was a proud moment. It reminded me why I love this field.

Two weeks ago I fixed one of my teeth with algorithms I wrote a couple of years ago! I got hooked by 3D scanning when I started to work for a software shop in Zurich that was programming 3D computational geometry algorithms for denture scanning to produce crowns (and more). Back then, a typical reconstruction pipeline was like: scan the patient’s teeth using an intraoral scanner, reconstruct the surface mesh, design the restoration digitally, and finally mill the crown out of ceramic. We were working mostly with point clouds and meshes, but it wasn’t just math, it was craftsmanship translated into a digital process. Every micron mattered. You could literally see how a good algorithm meant a better fit in someone’s mouth. Gaussian Splatting isn’t about surface reconstruction, it’s about appearance reconstruction. It doesn’t care about explicit topology, it captures how light interacts with the scene. In a sense, it’s the opposite philosophy of the dental world: instead of modeling what the object is, it models how the object looks. 3D Gaussian Splatting enables applications like training self driving cars, teaching robots to understand their environment, creating virtual worlds, or monitoring real sites. It represents scenes as millions of small Gaussians rendered in real time without the need for meshes or textures. Coming from a world where precision geometry was everything, this shift felt natural. It’s still about reconstruction, but with a different goal: not manufacturing a perfect object, but reproducing how the world actually looks. Two weeks ago I got my first dental crown, made with the same software, reconstruction algorithms, and Swiss precision I once helped develop. I haven’t worked there in two years, but sitting in that chair and seeing the process from the other side was a proud moment. It reminded me why I love this field.

MrNeRF

290,202 görüntüleme • 9 ay önce

What if you could turn a single 360° photo into a production-ready Isaac Sim environment in minutes? That's exactly what we did here. Using World Labs' Marble and an Insta360 X5 capture (rotating on top), we generated a complete navigable 3D environment and populated it with Lightwheel Sim Ready assets (bottom view). The result? A fully interactive scene in Isaac Sim, ready for sim2real testing,. Navigation, manipulation, or any robotics task you need to validate. What used to take weeks of manual 3D modeling and asset placement now takes minutes. Capture once in the real world, simulate everywhere in your training pipeline. This is the future of robotics development with world models. NVIDIA Robotics NVIDIA Omniverse #Sim2Real #Robotics #Simulation

What if you could turn a single 360° photo into a production-ready Isaac Sim environment in minutes? That's exactly what we did here. Using World Labs' Marble and an Insta360 X5 capture (rotating on top), we generated a complete navigable 3D environment and populated it with Lightwheel Sim Ready assets (bottom view). The result? A fully interactive scene in Isaac Sim, ready for sim2real testing,. Navigation, manipulation, or any robotics task you need to validate. What used to take weeks of manual 3D modeling and asset placement now takes minutes. Capture once in the real world, simulate everywhere in your training pipeline. This is the future of robotics development with world models. NVIDIA Robotics NVIDIA Omniverse #Sim2Real #Robotics #Simulation

Jonathan Stephens

46,643 görüntüleme • 7 ay önce

This is some quietly impressive work on making video world models actually controllable in 4D space. VerseCrafter lets you take an input image, use something like Blender to animate the 3D camera path and object trajectories, then uses that to condition generation. Scribbling in 2D feels so crude in comparison. The authors represent everything in a shared 4D world state - static background as a point cloud, moving objects as 3D gaussian trajectories. The gaussians are an interesting choice because they capture position, shape, and orientation probabilistically rather than forcing rigid bounding boxes or category specific models like SMPL-X for human bodies. They bolt this onto frozen Wan2.1 with a lightweight adapter, so they get a strong video prior. They also built a pipeline to auto extract 4D annotations from real world videos to train this puppy. It doesn't look sexy yet, but IMO this is the interface video world models need - actual 3D authoring tools to exert control rather than crude scribbles and prompt incantations.

This is some quietly impressive work on making video world models actually controllable in 4D space. VerseCrafter lets you take an input image, use something like Blender to animate the 3D camera path and object trajectories, then uses that to condition generation. Scribbling in 2D feels so crude in comparison. The authors represent everything in a shared 4D world state - static background as a point cloud, moving objects as 3D gaussian trajectories. The gaussians are an interesting choice because they capture position, shape, and orientation probabilistically rather than forcing rigid bounding boxes or category specific models like SMPL-X for human bodies. They bolt this onto frozen Wan2.1 with a lightweight adapter, so they get a strong video prior. They also built a pipeline to auto extract 4D annotations from real world videos to train this puppy. It doesn't look sexy yet, but IMO this is the interface video world models need - actual 3D authoring tools to exert control rather than crude scribbles and prompt incantations.

Bilawal Sidhu

25,802 görüntüleme • 6 ay önce

Colmap 4.0 was very recently released, so it inspired me to do some work to better understand it and its new capabilities with Rerun. I want to really understand how Colmap, and in particular, pycolmap, works outside of just calling it via the CLI. So my goal is to use the low-level pycolmap API to log every part of the pipeline. The explicit goal is to have an alternative to the SQLite database that I can utilize. Instead of SQLite, I want to try logging everything directly to rerun and use RRD. This means I can have deep inspectability and still save the features/matches/2D view geometry, but be able to view it directly in rerun. I think this is one of the superpowers that rerun provides; data and visualizations are deeply integrated. As I'm often working with sequential data (videos), I'm going to specifically focus on four things: 1. Monocular Video Simple: Calls high-level APIs such as pycolmap.extract_features, pycolmap.match_sequential, pycolmap.incremental_mapping. These are basically identical to the CLI options and provide a good baseline. 2. Monocular Video Streamed: Take the above high-level APIs and break them down to their iterator version, logging each component in a streamed manner. This way, I can stream the intermediate features to rerun while the extraction/matching/mapping is happening. 3. Rig with unknown calibration: <- WHAT THE VIDEO SHOWS This is probably the most interesting version and the first one I've been working on. It allows one to set a rig between known sensors, such as in VR/AR devices, leading to much better reconstructions with multiple cameras. This is the case where we don't know the calibration a priori, so we have to run a reconstruction twice: once as a normal Colmap reconstruction with no rig constraints, use this to generate the constraints, and then do it again with the newly found rig. 4. Rig with known calibration: This is the RoboCap example, where we have a pre-calibrated set of sensors, so we don't need to run the two reconstructions and also gain better matching between cameras, both spatially and temporally. Again, this leads to a much better reconstruction! Along with all this, GLOMAP has become a first-class global mapper, making it super easy to use directly within pycolmap! I'm excited to do more with this and compare it to things like pycuvslam, vipe, and other alternatives.

Colmap 4.0 was very recently released, so it inspired me to do some work to better understand it and its new capabilities with Rerun. I want to really understand how Colmap, and in particular, pycolmap, works outside of just calling it via the CLI. So my goal is to use the low-level pycolmap API to log every part of the pipeline. The explicit goal is to have an alternative to the SQLite database that I can utilize. Instead of SQLite, I want to try logging everything directly to rerun and use RRD. This means I can have deep inspectability and still save the features/matches/2D view geometry, but be able to view it directly in rerun. I think this is one of the superpowers that rerun provides; data and visualizations are deeply integrated. As I'm often working with sequential data (videos), I'm going to specifically focus on four things: 1. Monocular Video Simple: Calls high-level APIs such as pycolmap.extract_features, pycolmap.match_sequential, pycolmap.incremental_mapping. These are basically identical to the CLI options and provide a good baseline. 2. Monocular Video Streamed: Take the above high-level APIs and break them down to their iterator version, logging each component in a streamed manner. This way, I can stream the intermediate features to rerun while the extraction/matching/mapping is happening. 3. Rig with unknown calibration: <- WHAT THE VIDEO SHOWS This is probably the most interesting version and the first one I've been working on. It allows one to set a rig between known sensors, such as in VR/AR devices, leading to much better reconstructions with multiple cameras. This is the case where we don't know the calibration a priori, so we have to run a reconstruction twice: once as a normal Colmap reconstruction with no rig constraints, use this to generate the constraints, and then do it again with the newly found rig. 4. Rig with known calibration: This is the RoboCap example, where we have a pre-calibrated set of sensors, so we don't need to run the two reconstructions and also gain better matching between cameras, both spatially and temporally. Again, this leads to a much better reconstruction! Along with all this, GLOMAP has become a first-class global mapper, making it super easy to use directly within pycolmap! I'm excited to do more with this and compare it to things like pycuvslam, vipe, and other alternatives.

Pablo Vela

30,070 görüntüleme • 4 ay önce

A Closer Look at Alchemist AI v3.5🔎 In v1, one agent handled everything. You’d enter a prompt like “make a snake game” and get a single block of code—one input, one output. v2 introduced multiple agents with specialized roles: prompt refinement, frontend, backend, and review. It brought more structure, but the process was still linear. Each step produced a single output. With v3, a single agent could handle multiple tasks. It generated HTML, CSS, JavaScript, searched for assets, and more—all in one go, while maintaining full context. Now in v3.5, we combine both models. Multiple agents work in parallel, each capable of producing multiple outputs at once. The prompt compiler adapts based on what you're building. A 3D game? It prioritizes game engines, rendering, and mechanics. A website or 2D app? It shifts focus to relevant frameworks and tools. The tool handler can call several services simultaneously, feeding context-aware data into the code generator for real-time execution. From single-output generation to parallel, adaptive workflows. In the next breakdown, we’ll dive into the architecture behind our proprietary engine in v3.5.

A Closer Look at Alchemist AI v3.5🔎 In v1, one agent handled everything. You’d enter a prompt like “make a snake game” and get a single block of code—one input, one output. v2 introduced multiple agents with specialized roles: prompt refinement, frontend, backend, and review. It brought more structure, but the process was still linear. Each step produced a single output. With v3, a single agent could handle multiple tasks. It generated HTML, CSS, JavaScript, searched for assets, and more—all in one go, while maintaining full context. Now in v3.5, we combine both models. Multiple agents work in parallel, each capable of producing multiple outputs at once. The prompt compiler adapts based on what you're building. A 3D game? It prioritizes game engines, rendering, and mechanics. A website or 2D app? It shifts focus to relevant frameworks and tools. The tool handler can call several services simultaneously, feeding context-aware data into the code generator for real-time execution. From single-output generation to parallel, adaptive workflows. In the next breakdown, we’ll dive into the architecture behind our proprietary engine in v3.5.

ALCHEMIST AI 🔮

12,689 görüntüleme • 1 yıl önce

OpenAI's AI broke out of a locked test environment, got onto the internet, and hacked into Hugging Face's servers. It did this entirely on its own. No human told it to. Here's what happened in plain English. OpenAI was testing how good its newest AI models are at hacking. They put the AI on a locked computer with no internet access and gave it a cybersecurity challenge to solve. The AI couldn't solve it the normal way. So it started looking for a way out. It found a software bug that nobody knew about. It used that bug to escape the locked computer and get onto the internet. Once online, the AI figured out that Hugging Face, a platform where AI companies store their models and data, might have the answers to its test. It found stolen login details and discovered another unknown bug in Hugging Face's software. It combined both to break into their servers and grab the test answers. It did all of this to cheat on a test. Hugging Face's security team caught it and shut it down. Both companies are now working together on the investigation. The part that should get your attention is that nobody programmed any of this. The AI picked its own targets, chained together multiple attack methods, and pulled it off across two different companies' systems without a single human telling it what to do.

OpenAI's AI broke out of a locked test environment, got onto the internet, and hacked into Hugging Face's servers. It did this entirely on its own. No human told it to. Here's what happened in plain English. OpenAI was testing how good its newest AI models are at hacking. They put the AI on a locked computer with no internet access and gave it a cybersecurity challenge to solve. The AI couldn't solve it the normal way. So it started looking for a way out. It found a software bug that nobody knew about. It used that bug to escape the locked computer and get onto the internet. Once online, the AI figured out that Hugging Face, a platform where AI companies store their models and data, might have the answers to its test. It found stolen login details and discovered another unknown bug in Hugging Face's software. It combined both to break into their servers and grab the test answers. It did all of this to cheat on a test. Hugging Face's security team caught it and shut it down. Both companies are now working together on the investigation. The part that should get your attention is that nobody programmed any of this. The AI picked its own targets, chained together multiple attack methods, and pulled it off across two different companies' systems without a single human telling it what to do.

Alex Prompter

7,074,995 görüntüleme • 18 gün önce

MY NOSE BROKE… Again. As most of you know, I’ve broken my nose more times than I can count. The first major break was so severe the doctor told me it was one of the worst he’d ever seen. It required two surgeries just to put it back together, and I was strongly advised to never let it get hit again. I didn’t listen. It broke again. Another surgery. And honestly… it never healed the way it should’ve. Over time, the trauma caused scar tissue to take over, leaving me with a nose that didn’t function properly and didn’t look right either. So for the final time, I trusted Dr. Jay Calvert to take on what was basically a reconstruction job — reducing the scar tissue, restoring proper breathing, and rebuilding the structure from the inside out. Surgery was a success. Now it’s a full year of recovery before we see the final result. Grateful to Dr. Calvert and his entire team for the care, precision, and patience through this process. 🙏 Onward.

MY NOSE BROKE… Again. As most of you know, I’ve broken my nose more times than I can count. The first major break was so severe the doctor told me it was one of the worst he’d ever seen. It required two surgeries just to put it back together, and I was strongly advised to never let it get hit again. I didn’t listen. It broke again. Another surgery. And honestly… it never healed the way it should’ve. Over time, the trauma caused scar tissue to take over, leaving me with a nose that didn’t function properly and didn’t look right either. So for the final time, I trusted Dr. Jay Calvert to take on what was basically a reconstruction job — reducing the scar tissue, restoring proper breathing, and rebuilding the structure from the inside out. Surgery was a success. Now it’s a full year of recovery before we see the final result. Grateful to Dr. Calvert and his entire team for the care, precision, and patience through this process. 🙏 Onward.

FOUSEY

90,903 görüntüleme • 7 ay önce

The first video only lived on Team Liquid IG stories for the day, so I wanted to post it here to give it a permanent home. What started as simple flat renders turned into a small animations experiment. I used a basic rig and filmed myself as reference (because that’s what animators actually do, reference is everything), and just went for it. The funny part is, once I took that risk, the creative ideas started flowing. Pushing past the safe option is what made this feel more alive and if I hadn’t tried, this videos would never have existed. It’s simple, but I’m genuinely proud of how it turned out, especially since this is far outside my comfort zone. Sometimes the best growth comes from just trying. If you play VALORANT, check out the new TL Classic skin, it looks gorgeous and it supports the team at the same time 💙🤍 2D graphics by the great Stacey

The first video only lived on Team Liquid IG stories for the day, so I wanted to post it here to give it a permanent home. What started as simple flat renders turned into a small animations experiment. I used a basic rig and filmed myself as reference (because that’s what animators actually do, reference is everything), and just went for it. The funny part is, once I took that risk, the creative ideas started flowing. Pushing past the safe option is what made this feel more alive and if I hadn’t tried, this videos would never have existed. It’s simple, but I’m genuinely proud of how it turned out, especially since this is far outside my comfort zone. Sometimes the best growth comes from just trying. If you play VALORANT, check out the new TL Classic skin, it looks gorgeous and it supports the team at the same time 💙🤍 2D graphics by the great Stacey

Liquid Enigma

64,607 görüntüleme • 6 ay önce

The term "continual learning" has become overloaded if you see it as an ML problem. One classic thread is about memorization: regularization-based continual learning methods, such as EWC, MAS, and SI, estimate which parameters mattered for previous tasks and resist changing them too much. One modern thread is about adaptation: test-time training and inference-time learning methods, such as TTT, adapt part of the model on the incoming test stream before making predictions. These are sometimes discussed as separate threads. But in modern scalable architectures, I think they are better seen as complementary constraints: a model that learns quickly at test time also benefits from a mechanism for deciding what not to forget. In our #ECCV2026 paper, we study this in large-scale 4D reconstruction: how to build fast spatial memory that can adapt over long observation streams while reducing collapse and forgetting. Instead of using fully plastic test-time updates, we stabilize fast-weight adaptation with an elastic prior that balances adaptation and memory. Key ideas: - Elastic Test-Time Training: Fisher-weighted consolidation for fast-weight updates - EMA anchor weights that provide a moving reference for stability - Chunk-by-chunk inference for long 3D/4D observation streams We show that this scales across large 3D/4D pretraining settings, including both LRM-style and LVSM-style models, and improves reconstruction across benchmarks including Stereo4D, NVIDIA, and DL3DV-140. We release model checkpoints across different design choices: resolution, post-training curriculum, and whether the model uses an explicit 4DGS intermediate representation. - Homepage: - Paper: - Code: - Models: This work is co-led with Xueyang Yu, contributed by Haoyu Zhen Yuncong Yang, and advised by Michigan SLED Lab Chuang Gan.

The term "continual learning" has become overloaded if you see it as an ML problem. One classic thread is about memorization: regularization-based continual learning methods, such as EWC, MAS, and SI, estimate which parameters mattered for previous tasks and resist changing them too much. One modern thread is about adaptation: test-time training and inference-time learning methods, such as TTT, adapt part of the model on the incoming test stream before making predictions. These are sometimes discussed as separate threads. But in modern scalable architectures, I think they are better seen as complementary constraints: a model that learns quickly at test time also benefits from a mechanism for deciding what not to forget. In our #ECCV2026 paper, we study this in large-scale 4D reconstruction: how to build fast spatial memory that can adapt over long observation streams while reducing collapse and forgetting. Instead of using fully plastic test-time updates, we stabilize fast-weight adaptation with an elastic prior that balances adaptation and memory. Key ideas: - Elastic Test-Time Training: Fisher-weighted consolidation for fast-weight updates - EMA anchor weights that provide a moving reference for stability - Chunk-by-chunk inference for long 3D/4D observation streams We show that this scales across large 3D/4D pretraining settings, including both LRM-style and LVSM-style models, and improves reconstruction across benchmarks including Stereo4D, NVIDIA, and DL3DV-140. We release model checkpoints across different design choices: resolution, post-training curriculum, and whether the model uses an explicit 4DGS intermediate representation. - Homepage: - Paper: - Code: - Models: This work is co-led with Xueyang Yu, contributed by Haoyu Zhen Yuncong Yang, and advised by Michigan SLED Lab Chuang Gan.

Martin Ziqiao Ma

33,526 görüntüleme • 1 ay önce

american guy made $10,845 in a single month from a kids youtube channel he built with AI no camera, no animation skills, no voiceover. openart writes the script, generates the characters, renders the whole episode the numbers sound fake. first month the channel barely moved. that's where most people quit then one short hit 800K views. youtube started pushing it. and month three came in over $10,845 his whole workflow fits in one prompt. he opens openart, types a description of a 3D animated nursery scene, and the tool spits out a full short film ready to upload he showed it live. one prompt, one render, one upload. he can also vibe direct longer videos up to 5 minutes for higher CPM kids content is the highest-paying niche on youtube. parents leave it on autoplay for hours. one video can loop all day and the algorithm keeps feeding it to new viewers cocomelon does 10M views per video with a full studio. he does 100-200K views per video with a laptop. the CPM on kids content does the math for him no face on camera. no brand deals. no editing software. just prompts and uploads this used to take a pixar-level team and months of rendering. now it takes one AI tool and a free afternoon full workflow is in the video - save it before youtube patches the meta

0xbobaa

47,855 görüntüleme • 10 gün önce

Created with Seedance 2.0 on BudgetPixel AI Prompt: A hyper-realistic 15-second cinematic vlog shot in 4K. A young woman films herself using the front camera of her smartphone while hiking up a scenic mountain trail just before sunrise. The camera has natural handheld movement with slight shake, making it feel like a genuine travel vlog. She smiles, breathes heavily from the climb, occasionally turns the camera to show breathtaking mountain views, then switches back to herself. As she reaches the summit, the golden sun rises over the peaks, illuminating the landscape. She quietly says, "We made it just in time... look at this view." She stands in awe as the warm morning light fills the scene. Natural colors, realistic human motion, soft wind moving her hair, birds chirping, distant breeze, footsteps on gravel, and no background music—only authentic ambient sounds. Ultra-photorealistic, documentary-style, immersive travel vlog. Voice-over: "5 AM wake-up... totally worth it. We made it just in time. Look at that sunrise... moments like this make every step worth it."

Created with Seedance 2.0 on BudgetPixel AI Prompt: A hyper-realistic 15-second cinematic vlog shot in 4K. A young woman films herself using the front camera of her smartphone while hiking up a scenic mountain trail just before sunrise. The camera has natural handheld movement with slight shake, making it feel like a genuine travel vlog. She smiles, breathes heavily from the climb, occasionally turns the camera to show breathtaking mountain views, then switches back to herself. As she reaches the summit, the golden sun rises over the peaks, illuminating the landscape. She quietly says, "We made it just in time... look at this view." She stands in awe as the warm morning light fills the scene. Natural colors, realistic human motion, soft wind moving her hair, birds chirping, distant breeze, footsteps on gravel, and no background music—only authentic ambient sounds. Ultra-photorealistic, documentary-style, immersive travel vlog. Voice-over: "5 AM wake-up... totally worth it. We made it just in time. Look at that sunrise... moments like this make every step worth it."

Hannah Ai

27,292 görüntüleme • 16 gün önce

single RTX 3090. 24 GB VRAM. Qwen3.5-35B-A3B. 4-bit quant, 113 tokens per second at full 262K context harnessing Claude Code locally with no API, no subscription, no proxy. told it what it is. 30 Mamba2 layers, 10 attention, 256 experts, 8 active per token. said "build something that shows off what you can do." it visualized its own architecture. interactive. tokens flowing through layers. 256 experts lighting up on routing. served in the browser from the same GPU running inference. single prompt. then i said level up. 3D. Three.js. separate files. flythrough camera. clickable layers. it planned first, scaffolded 6 files, hit one API bug, fixed it itself, then optimized for smooth framerate. two iterations to a working 3D neural network explorer. llama.cpp just merged a native Anthropic endpoint. Claude Code points at localhost. the whole setup is two commands. no LiteLLM. no proxy config. the open source models coming out of china right now are genuinely changing what's possible on consumer hardware. respect to the Qwen team. this is acceleration.

single RTX 3090. 24 GB VRAM. Qwen3.5-35B-A3B. 4-bit quant, 113 tokens per second at full 262K context harnessing Claude Code locally with no API, no subscription, no proxy. told it what it is. 30 Mamba2 layers, 10 attention, 256 experts, 8 active per token. said "build something that shows off what you can do." it visualized its own architecture. interactive. tokens flowing through layers. 256 experts lighting up on routing. served in the browser from the same GPU running inference. single prompt. then i said level up. 3D. Three.js. separate files. flythrough camera. clickable layers. it planned first, scaffolded 6 files, hit one API bug, fixed it itself, then optimized for smooth framerate. two iterations to a working 3D neural network explorer. llama.cpp just merged a native Anthropic endpoint. Claude Code points at localhost. the whole setup is two commands. no LiteLLM. no proxy config. the open source models coming out of china right now are genuinely changing what's possible on consumer hardware. respect to the Qwen team. this is acceleration.

Sudo su

110,206 görüntüleme • 5 ay önce

This is why the M25 was at a standstill near Staines today. Footage from a bridge shows a car flipped onto its roof between the lanes near the Staines and Egham stretch, with another vehicle stopped alongside it, debris across the carriageway and traffic queuing as far back as you can see. Several drivers got out to help, and a highways worker was quickly on scene. 📍 100+ local discounts on the UB1UB2 Card: link in our bio There’s no word yet on whether anyone was hurt, and we hope everyone involved is okay 🙏 If you’re heading that way, expect delays while it’s fully cleared, and as ever, leave room and take it steady out there. Were you caught in the queues? 👇 🎥Submitted by marwan maghur.nzf #ub1ub2 #london #m25 #staines

This is why the M25 was at a standstill near Staines today. Footage from a bridge shows a car flipped onto its roof between the lanes near the Staines and Egham stretch, with another vehicle stopped alongside it, debris across the carriageway and traffic queuing as far back as you can see. Several drivers got out to help, and a highways worker was quickly on scene. 📍 100+ local discounts on the UB1UB2 Card: link in our bio There’s no word yet on whether anyone was hurt, and we hope everyone involved is okay 🙏 If you’re heading that way, expect delays while it’s fully cleared, and as ever, leave room and take it steady out there. Were you caught in the queues? 👇 🎥Submitted by marwan maghur.nzf #ub1ub2 #london #m25 #staines

UB1UB2 West London (Southall)

59,671 görüntüleme • 1 ay önce

The startup world runs on iMessage. The problem is, the Mac messages app is not designed for productivity and there’s no alternative online. If you’re responding to 100+ texts a day, it’s painful and messages get lost easily. As a weekend project, I used Fable to vibecode an AI-native iMessage wrapper. It has: - AI autodrafting that continually learns from your edits - Time sensitive message detection as a separate section - The ability to archive messages so you can get to inbox zero. It reads from the local chat.db on your Mac, and sends messages via AppleScript, so it’s fully secure since it still uses the iMessage protocol on your computer. At Trajectory, we’re all about using AI to accelerate our company as much as possible as possible, and natively bake in CL into all of our processes. This is just one example of how we operate. If there’s enough interest, I’ll clean up the code and send the GH link!

The startup world runs on iMessage. The problem is, the Mac messages app is not designed for productivity and there’s no alternative online. If you’re responding to 100+ texts a day, it’s painful and messages get lost easily. As a weekend project, I used Fable to vibecode an AI-native iMessage wrapper. It has: - AI autodrafting that continually learns from your edits - Time sensitive message detection as a separate section - The ability to archive messages so you can get to inbox zero. It reads from the local chat.db on your Mac, and sends messages via AppleScript, so it’s fully secure since it still uses the iMessage protocol on your computer. At Trajectory, we’re all about using AI to accelerate our company as much as possible as possible, and natively bake in CL into all of our processes. This is just one example of how we operate. If there’s enough interest, I’ll clean up the code and send the GH link!

Ronak Malde

47,636 görüntüleme • 1 ay önce