Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

Cool new model release: PaintsUndo, which takes an image and outputs step-by-step frames of how to draw it. It starts with a sketch and then fills in the colors, shading, and lighting.

venturetwins's profile picture

Justine Moore

181,219 subscribers

146,126 Aufrufe • vor 1 Jahr •via X (Twitter)

Wissenschaft & TechnologieBildungKunst
LIVE

Anya Rossi• Live Now

Private livecam show

10 Kommentare

Profilbild von Justine Moore
Justine Moorevor 1 Jahr

It runs locally and takes 5-10 minutes to produce a ~25 second transformation video, with 4 fps. You an also use it to extract single frames of sketches.

Profilbild von Justine Moore
Justine Moorevor 1 Jahr

Here's the project page with a bunch more examples (it also links to the GitHub). This model was made by @lvminzhang, who continues to put out 🔥 work in the Stable Diffusion ecosystem.

Profilbild von Justine Moore
Justine Moorevor 1 Jahr

Okay updating this to add that you can now run it in a @huggingface Space! (as a side note, it’s fun to see which spaces on ZeroGPU are most popular…I expect this will be in the top soon)

Profilbild von George Crudo
George Crudovor 1 Jahr

there is no good intentioned value to having an AI construct a fictional creation process

Profilbild von Justine Moore
Justine Moorevor 1 Jahr

Oh I disagree, I think people will find plenty of interesting things to do with this! Starting with learning how to draw by following much clearer steps (replacing this)

Profilbild von Benjamin BLM
Benjamin BLMvor 1 Jahr

The primary utility of this "tool" is scamming people. Currently, artists are posting process videos to show their work and prove they're not using AI. This tool seems directly targeted on destroying peoples ability to determine if artwork was human made.

Profilbild von Wetterschneider
Wetterschneidervor 1 Jahr

Mimicry of process is worse than mimicry of final result. This is a tool for ripping people off. That's its only function. Why are you promoting this? And btw, it's non-functional mimicry. A person showing off this garbage is going to leave a trail of disappointed "clients".

Profilbild von Benjamin BLM
Benjamin BLMvor 1 Jahr

I'm willing to bet that you have never actually picked up a "how to draw x" book, or looked at an online drawing tutorial, because this is not even remotely useful. This isn't a step-by-step how to draw. It's a bastardization of timelapses, faking a process.

Profilbild von bitbutter
bitbuttervor 1 Jahr

I'm not seeing any real construction lines. so it's unclear that this has value as a training aid in its current state.

Profilbild von 🦋
🦋vor 1 Jahr

Thank you for sharing...this is great evidence for all the court cases, Congress, FTC and Parliament to realise that generative AI is detrimental & Criminal to our society & future. When fraud & exploitation are the soul & intention of Generative AI...it has no place in society.

Ähnliche Videos

Step-by-step Lion Sketch Tutorial! From simple lines to a powerful king of the jungle 👑 — learn how to draw a lion with just a pencil.
1:19

Step-by-step Lion Sketch Tutorial! From simple lines to a powerful king of the jungle 👑 — learn how to draw a lion with just a pencil.

Living Tricks

16,524 Aufrufe • vor 8 Monaten

Introducing Frames: An image generation model offering unprecedented stylistic control. Frames is our newest foundation model for image generation, marking a big step forward in stylistic control and visual fidelity. With Frames, you can begin to architect worlds that represent very specific points of view and aesthetic characteristics. See below for examples. World 1089: Mise-en-scène (1/11)
0:36

Introducing Frames: An image generation model offering unprecedented stylistic control. Frames is our newest foundation model for image generation, marking a big step forward in stylistic control and visual fidelity. With Frames, you can begin to architect worlds that represent very specific points of view and aesthetic characteristics. See below for examples. World 1089: Mise-en-scène (1/11)

Runway

263,304 Aufrufe • vor 1 Jahr

Grok Imagine lets you steal the style of any image: colors, lighting, texture... You can tweak it, remix it, and then generate an entire video that holds that same look across totally new scenes!
1:04

Grok Imagine lets you steal the style of any image: colors, lighting, texture... You can tweak it, remix it, and then generate an entire video that holds that same look across totally new scenes!

Mario Nawfal

16,329 Aufrufe • vor 4 Monaten

GPT Image 2 + Codex: or how to make Codex not suck at UI. Step 1: Generate a UI image (native in Codex) Step 2: Get Codex to implement the UI based on it Step 3: Get Codex to iterate until it aligns with the image as much as possible Codex is bad at initial UI, but very good at implementing a reference design, so this is your way out - iterate with the image model first and then Codex will do a good job.
0:37

GPT Image 2 + Codex: or how to make Codex not suck at UI. Step 1: Generate a UI image (native in Codex) Step 2: Get Codex to implement the UI based on it Step 3: Get Codex to iterate until it aligns with the image as much as possible Codex is bad at initial UI, but very good at implementing a reference design, so this is your way out - iterate with the image model first and then Codex will do a good job.

Peter Gostev

66,675 Aufrufe • vor 2 Monaten

Sometimes the best way to communicate your vision is to draw it. ✍️ Watch how we used a rough sketch to lock-in the composition, Nano Banana Pro to refine the look, and then Frames to Video to bring it all together. #FindYourFlow
0:38

Sometimes the best way to communicate your vision is to draw it. ✍️ Watch how we used a rough sketch to lock-in the composition, Nano Banana Pro to refine the look, and then Frames to Video to bring it all together. #FindYourFlow

Google Flow

27,735 Aufrufe • vor 5 Monaten

Using a pasta machine to send a photo through creates pixelated, simplified versions of the image. This is a great way to illustrate how Convolutional Neural Networks (CNNs) work. Each pass through the machine represents a layer in a CNN, where images are processed step by step. In CNNs, each layer identifies specific features like edges, textures, or colors. Just as the pasta machine alters the photo with each pass, CNNs refine and simplify the image for better understanding. This step-by-step process allows CNNs to recognize complex images, making them useful in areas like image classification and object detection. It ain’t perfect but it’s a simple presentation.
0:43

Using a pasta machine to send a photo through creates pixelated, simplified versions of the image. This is a great way to illustrate how Convolutional Neural Networks (CNNs) work. Each pass through the machine represents a layer in a CNN, where images are processed step by step. In CNNs, each layer identifies specific features like edges, textures, or colors. Just as the pasta machine alters the photo with each pass, CNNs refine and simplify the image for better understanding. This step-by-step process allows CNNs to recognize complex images, making them useful in areas like image classification and object detection. It ain’t perfect but it’s a simple presentation.

Brian Roemmele

97,202 Aufrufe • vor 1 Jahr

How to draw an Espeon: Step 1, draw a Jolteon Step 2, finish the rest of the damn Espeon
2:03

How to draw an Espeon: Step 1, draw a Jolteon Step 2, finish the rest of the damn Espeon

🎨Sifyro

52,376 Aufrufe • vor 5 Monaten

A new real-time world model is here 👀 I tested out Dynamics Lab's Mirage 2, which was just publicly released. You can upload an image and step inside it - with game controls that guide movement + the ability to change the scene with text prompts.
0:25

A new real-time world model is here 👀 I tested out Dynamics Lab's Mirage 2, which was just publicly released. You can upload an image and step inside it - with game controls that guide movement + the ability to change the scene with text prompts.

Justine Moore

24,297 Aufrufe • vor 10 Monaten

STOP using the crossover step & START using a drop step! The crossover step is slower, it starts the athlete in a horrible position for acceleration, and takes away from the athletes natural athleticism.
0:24

STOP using the crossover step & START using a drop step! The crossover step is slower, it starts the athlete in a horrible position for acceleration, and takes away from the athletes natural athleticism.

Alex Simone

43,641 Aufrufe • vor 3 Monaten

How to Make an Image Transparent (Step by Step Guide) ( pro_graphic7 )
0:34

How to Make an Image Transparent (Step by Step Guide) ( pro_graphic7 )

Mutia

29,927 Aufrufe • vor 9 Monaten

NEW EPISODE: “Images have frames or borders because it's like you enter it, and it's not a real thing. It's an image. Then in the image, everything is possible.” -Urs Fischer
1:44:46

NEW EPISODE: “Images have frames or borders because it's like you enter it, and it's not a real thing. It's an image. Then in the image, everything is possible.” -Urs Fischer

Tetragrammaton

55,922 Aufrufe • vor 11 Monaten

Vidu is the craziest AI video model I’ve tested this week. I used just a prompt and one image, and got cinematic results with motion, lighting, and effects. Hard to believe this level of cinematic videos it can generate. 10 wild outputs: 1. Crime thriller
0:23

Vidu is the craziest AI video model I’ve tested this week. I used just a prompt and one image, and got cinematic results with motion, lighting, and effects. Hard to believe this level of cinematic videos it can generate. 10 wild outputs: 1. Crime thriller

Atal

40,114 Aufrufe • vor 1 Jahr

Announcing Ideogram Editor! Now you can draw an image and control the color and composition of Ideogram outputs. You can also edit typography images with more precision and crop pictures with new aspect ratios!
1:11

Announcing Ideogram Editor! Now you can draw an image and control the color and composition of Ideogram outputs. You can also edit typography images with more precision and crop pictures with new aspect ratios!

Ideogram

73,020 Aufrufe • vor 2 Jahren

How to draw a man's arm step by step
0:21

How to draw a man's arm step by step

Next Step ©

52,921 Aufrufe • vor 1 Monat

OpenAI just released the number 1 AI opportunity of 2025 The ChatGPT App Store was just announced, and you need to be building for it immediately In this video I cover how their new app store works, and walk you step by step how to build an app for it (no coding exp required):
12:32

OpenAI just released the number 1 AI opportunity of 2025 The ChatGPT App Store was just announced, and you need to be building for it immediately In this video I cover how their new app store works, and walk you step by step how to build an app for it (no coding exp required):

Alex Finn

123,963 Aufrufe • vor 8 Monaten

From an artist's mind, to the game: every detail tells a story. Episode 3 of From The Ward showcases the step by step process of how Joel and the team at Infinity Ward take iconic characters and gear from rough sketch, to your loadout in Call of Duty: Modern Warfare 4.
9:04

From an artist's mind, to the game: every detail tells a story. Episode 3 of From The Ward showcases the step by step process of how Joel and the team at Infinity Ward take iconic characters and gear from rough sketch, to your loadout in Call of Duty: Modern Warfare 4.

Call of Duty

207,872 Aufrufe • vor 21 Tagen

TurboEdit Instant text-based image editing discuss: We address the challenges of precise image inversion and disentangled image editing in the context of few-step diffusion models. We introduce an encoder based iterative inversion technique. The inversion network is conditioned on the input image and the reconstructed image from the previous step, allowing for correction of the next reconstruction towards the input image. We demonstrate that disentangled controls can be easily achieved in the few-step diffusion model by conditioning on an (automatically generated) detailed text prompt. To manipulate the inverted image, we freeze the noise maps and modify one attribute in the text prompt (either manually or via instruction based editing driven by an LLM), resulting in the generation of a new image similar to the input image with only one attribute changed. It can further control the editing strength and accept instructive text prompt. Our approach facilitates realistic text-guided image edits in real-time, requiring only 8 number of functional evaluations (NFEs) in inversion (one-time cost) and 4 NFEs per edit. Our method is not only fast, but also significantly outperforms state-of-the-art multi-step diffusion editing techniques.
0:22

TurboEdit Instant text-based image editing discuss: We address the challenges of precise image inversion and disentangled image editing in the context of few-step diffusion models. We introduce an encoder based iterative inversion technique. The inversion network is conditioned on the input image and the reconstructed image from the previous step, allowing for correction of the next reconstruction towards the input image. We demonstrate that disentangled controls can be easily achieved in the few-step diffusion model by conditioning on an (automatically generated) detailed text prompt. To manipulate the inverted image, we freeze the noise maps and modify one attribute in the text prompt (either manually or via instruction based editing driven by an LLM), resulting in the generation of a new image similar to the input image with only one attribute changed. It can further control the editing strength and accept instructive text prompt. Our approach facilitates realistic text-guided image edits in real-time, requiring only 8 number of functional evaluations (NFEs) in inversion (one-time cost) and 4 NFEs per edit. Our method is not only fast, but also significantly outperforms state-of-the-art multi-step diffusion editing techniques.

AK

16,062 Aufrufe • vor 1 Jahr