Video yükleniyor...

Video Yüklenemedi

Ana Sayfaya Dön

Flow matching produces smooth, deterministic trajectories. In contrast, the sampling process of a diffusion model is chaotic, resembling the random motion of gas particles.

alec_helbling's profile picture

Alec Helbling

11,704 subscribers

437,025 görüntüleme • 1 yıl önce •via X (Twitter)

SanatBilim & TeknolojiEğitim
LIVE

Anya Rossi• Live Now

Private livecam show

0 Yorum

Yorum bulunmuyor

Orijinal gönderinin yorumları burada görünecek

Benzer Videolar

Diffusion models leverage a variety of samplers. Deterministic methods like DDIM produce orderly paths. In contrast, stochastic samplers like DDPM produce chaotic trajectories. Despite their differences, both methods draw valid samples from the underlying distribution.
0:10

Diffusion models leverage a variety of samplers. Deterministic methods like DDIM produce orderly paths. In contrast, stochastic samplers like DDPM produce chaotic trajectories. Despite their differences, both methods draw valid samples from the underlying distribution.

Alec Helbling

103,040 görüntüleme • 1 yıl önce

Stochastic and deterministic sampling strategies for diffusion models produce strikingly different trajectories, but both ultimately achieve the same aim. I had a great time presenting our work, Diffusion Explorer, this week at IEEE VIS in Vienna.
0:14

Stochastic and deterministic sampling strategies for diffusion models produce strikingly different trajectories, but both ultimately achieve the same aim. I had a great time presenting our work, Diffusion Explorer, this week at IEEE VIS in Vienna.

Alec Helbling

279,334 görüntüleme • 10 ay önce

The Cucker-Smale system of ODEs is the simplest model of particles exhibiting flocking behaviors. Diffusion of speeds is governed by a Laplacian on the fully connected graph of points.
0:10

The Cucker-Smale system of ODEs is the simplest model of particles exhibiting flocking behaviors. Diffusion of speeds is governed by a Laplacian on the fully connected graph of points.

Gabriel Peyré

77,155 görüntüleme • 2 yıl önce

In flow matching, a coupling determines how noise and data samples are paired during training. The choice of coupling is important because it influences the geometry of trajectories at inference time. The simplest choice is the independent coupling, where noise and data points are paired arbitrarily. This can lead to curved trajectories as the model averages over many conflicting pairings. However, if we use optimal transport on batches of pairs, this leads to fewer ambiguous intersections that the model must resolve, leading to straighter trajectories at inference time.
0:18

In flow matching, a coupling determines how noise and data samples are paired during training. The choice of coupling is important because it influences the geometry of trajectories at inference time. The simplest choice is the independent coupling, where noise and data points are paired arbitrarily. This can lead to curved trajectories as the model averages over many conflicting pairings. However, if we use optimal transport on batches of pairs, this leads to fewer ambiguous intersections that the model must resolve, leading to straighter trajectories at inference time.

Alec Helbling

65,484 görüntüleme • 4 ay önce

Oldies but goldies: Felipe Cucker, Steve Smale, The Mathematics of Emergence, 2007. Defines a simple model of particles exhibiting flocking behaviors. Diffusion of speeds is governed by a Laplacian on the fully connected graph of points.
0:10

Oldies but goldies: Felipe Cucker, Steve Smale, The Mathematics of Emergence, 2007. Defines a simple model of particles exhibiting flocking behaviors. Diffusion of speeds is governed by a Laplacian on the fully connected graph of points.

Gabriel Peyré

24,386 görüntüleme • 2 yıl önce

Do you enjoy the process of carving wood shapes that produces very smooth and precise shapes?
0:14

Do you enjoy the process of carving wood shapes that produces very smooth and precise shapes?

Calisa Natanya

351,723 görüntüleme • 4 ay önce

We have released the code and weights for our #CVPR2023 paper "Avatars Grow Legs: Generating Smooth Human Motion from Sparse Tracking Inputs with Diffusion Model"! code: abs: project: The demo is below:
0:14

We have released the code and weights for our #CVPR2023 paper "Avatars Grow Legs: Generating Smooth Human Motion from Sparse Tracking Inputs with Diffusion Model"! code: abs: project: The demo is below:

Artsiom Sanakoyeu

35,715 görüntüleme • 3 yıl önce

The Dream 7B (diffusion reasoning language model) is OUT! 🚨 I built a demo so you can test it out (and check the diffusion process live) 𖣯🔍
0:10

The Dream 7B (diffusion reasoning language model) is OUT! 🚨 I built a demo so you can test it out (and check the diffusion process live) 𖣯🔍

apolinario (poli)

35,679 görüntüleme • 1 yıl önce

I've been putting together an interactive tool called DiffusionLab for explaining the geometric intuition behind diffusion and flow based generative models. Sampling is actually being done in the browser using Tensorflow.js! It is still in the very early stages.
0:11

I've been putting together an interactive tool called DiffusionLab for explaining the geometric intuition behind diffusion and flow based generative models. Sampling is actually being done in the browser using Tensorflow.js! It is still in the very early stages.

Alec Helbling

104,707 görüntüleme • 1 yıl önce

LLaDA (the first Large Language Diffusion Model) is *just* out 💥 and I've built a demo, try out now 👨‍💻 It's mesmerizing to watch the diffusion process 🌀, and it being a diffusion model gives you superpowers like "the 4th word has to be pineapple" 🦸 Demo and weights 👇
0:17

LLaDA (the first Large Language Diffusion Model) is *just* out 💥 and I've built a demo, try out now 👨‍💻 It's mesmerizing to watch the diffusion process 🌀, and it being a diffusion model gives you superpowers like "the 4th word has to be pineapple" 🦸 Demo and weights 👇

apolinario (poli)

82,599 görüntüleme • 1 yıl önce

Random Circuit Sampling (RCS) is a necessary benchmark to measure quantum computer performance in the presence of noise. Why is RCS considered a cornerstone of quantum computing? Find out in our latest breakdown. #QuantumAI
0:20

Random Circuit Sampling (RCS) is a necessary benchmark to measure quantum computer performance in the presence of noise. Why is RCS considered a cornerstone of quantum computing? Find out in our latest breakdown. #QuantumAI

Google Quantum AI

11,807 görüntüleme • 1 yıl önce

Lorenz attractor is the set of limit trajectories for a simplified mathematical model for atmospheric convection. It has a fractal geometry.
0:15

Lorenz attractor is the set of limit trajectories for a simplified mathematical model for atmospheric convection. It has a fractal geometry.

Gabriel Peyré

168,783 görüntüleme • 3 yıl önce

Super excited about Hydra-0 from Hongyu Li and team! The key idea is to use flow as a shared visual interface across embodiments/objects for controllable video generation, allowing a single generalist world model to learn from human, handheld-gripper, and robot interaction data. My favorite result is the video below: start from a real video of a human doing the task (left), extract the desired object flow, and condition the model on that flow (right). The model then hallucinates a plausible robot motion that could produce the same object motion. Very cool glimpse of how a generalist world model can bridge human demonstrations and robot control.
0:13

Super excited about Hydra-0 from Hongyu Li and team! The key idea is to use flow as a shared visual interface across embodiments/objects for controllable video generation, allowing a single generalist world model to learn from human, handheld-gripper, and robot interaction data. My favorite result is the video below: start from a real video of a human doing the task (left), extract the desired object flow, and condition the model on that flow (right). The model then hallucinates a plausible robot motion that could produce the same object motion. Very cool glimpse of how a generalist world model can bridge human demonstrations and robot control.

Yunzhu Li

10,695 görüntüleme • 27 gün önce

I just spent weeks handwriting a 200+ page guide on the entirety of Robotics foundations from scratch. Featuring: - Kinematics (Forward, Inverse Kin, Jacobians, DoF) - Controls (PID, Force/Impedance Control, Feedforward) - SLAM (Bayes Filters, Kalman, Particle Filters, Mapping) - Motion Planning (Dijkstra, A*, RRT, RRT*) - RL (MDP, Policy Gradients, PPO, Sim2Real, Value Func) - VLA (Q/K/V, ViT, SigLIP, Diffusion/Flow-Matching) - ACT, Diffusion Policy, UMI, LeRobot tooling - World Model (Dreamer, VideoGen, Gaussian Splatting) Everything to get started, for free.
0:19

I just spent weeks handwriting a 200+ page guide on the entirety of Robotics foundations from scratch. Featuring: - Kinematics (Forward, Inverse Kin, Jacobians, DoF) - Controls (PID, Force/Impedance Control, Feedforward) - SLAM (Bayes Filters, Kalman, Particle Filters, Mapping) - Motion Planning (Dijkstra, A*, RRT, RRT*) - RL (MDP, Policy Gradients, PPO, Sim2Real, Value Func) - VLA (Q/K/V, ViT, SigLIP, Diffusion/Flow-Matching) - ACT, Diffusion Policy, UMI, LeRobot tooling - World Model (Dreamer, VideoGen, Gaussian Splatting) Everything to get started, for free.

Arjun Virk

65,180 görüntüleme • 2 ay önce

We are also releasing self-contained lecture notes that explain flow matching and diffusion models from scratch. This goes from "zero" to the state-of-the-art in modern Generative AI. 📖 Read the notes here: Joint work with Ezra Erives.
0:15

We are also releasing self-contained lecture notes that explain flow matching and diffusion models from scratch. This goes from "zero" to the state-of-the-art in modern Generative AI. 📖 Read the notes here: Joint work with Ezra Erives.

Peter Holderrieth

476,548 görüntüleme • 6 ay önce

(1/2) Check out "𝐏𝐨𝐥𝐲𝐃𝐢𝐟𝐟: Generating 3D Polygonal Meshes with Diffusion Models"! Our model operates directly on the polygons of 3D meshes and generates novel shapes as output through an iterative diffusion process.
0:11

(1/2) Check out "𝐏𝐨𝐥𝐲𝐃𝐢𝐟𝐟: Generating 3D Polygonal Meshes with Diffusion Models"! Our model operates directly on the polygons of 3D meshes and generates novel shapes as output through an iterative diffusion process.

Matthias Niessner

57,947 görüntüleme • 2 yıl önce

a visualization of the flow field of the navier-stokes finite-time blowup solution. it's a good model sir (singularity at t = 1 with viscosity matching water, unit distance ~ 1 mm. leading and first order background terms)
0:18

a visualization of the flow field of the navier-stokes finite-time blowup solution. it's a good model sir (singularity at t = 1 with viscosity matching water, unit distance ~ 1 mm. leading and first order background terms)

penlu

170,095 görüntüleme • 8 gün önce

The fine line between smooth motion and engine damage is just a single strand of human hair
0:17

The fine line between smooth motion and engine damage is just a single strand of human hair

Interesting AF

975,763 görüntüleme • 11 ay önce

🚀New paper out - We present Video-MSG (Multimodal Sketch Guidance), a novel planning-based training-free guidance method for T2V models, improving control of spatial layout and object trajectories. 🔧 Key idea: • Generate a Video Sketch — a spatio-temporal plan with background, foreground, and motion in the pixel space. • Encode this structure directly into the latent space of the diffusion model during generation, which does not require fine-tuning or additional memory during inference. 🧵
0:12

🚀New paper out - We present Video-MSG (Multimodal Sketch Guidance), a novel planning-based training-free guidance method for T2V models, improving control of spatial layout and object trajectories. 🔧 Key idea: • Generate a Video Sketch — a spatio-temporal plan with background, foreground, and motion in the pixel space. • Encode this structure directly into the latent space of the diffusion model during generation, which does not require fine-tuning or additional memory during inference. 🧵

Jialu Li

35,258 görüntüleme • 1 yıl önce