Video yükleniyor...

Video Yüklenemedi

Ana Sayfaya Dön

I've been experimenting with embeddings a lot recently. Here's an interface I created that lets me interpolate between CLIP image embeddings and generate images from the intermediate latents Latent Surfing. it's kinda like a color picker, but for images. (1/n)

gd3kr's profile picture

aarya

12,650 subscribers

122,731 görüntüleme • 2 yıl önce •via X (Twitter)

EğitimBilim & Teknoloji
LIVE

Anya Rossi• Live Now

Private livecam show

0 Yorum

Yorum bulunmuyor

Orijinal gönderinin yorumları burada görünecek

Benzer Videolar

Create a 3D model from a single image, set of images or a text prompt in < 1 minute 😮‍💨 This new AI paper called CAT3D shows us that it’ll keep getting easier to produce 3D models from 2D images — whether it’s a sparser real world 3D scan (a few photos instead of hundreds) or your favorite 2D image generator like Midjourney (just an image). How does this magic work? “This architecture is similar to video diffusion models, but with camera pose embeddings for each image instead of time embeddings. The generated views are passed into a robust 3D reconstruction pipeline to create the 3D representation (Zip-NeRF or 3DGS)”
0:18

Create a 3D model from a single image, set of images or a text prompt in < 1 minute 😮‍💨 This new AI paper called CAT3D shows us that it’ll keep getting easier to produce 3D models from 2D images — whether it’s a sparser real world 3D scan (a few photos instead of hundreds) or your favorite 2D image generator like Midjourney (just an image). How does this magic work? “This architecture is similar to video diffusion models, but with camera pose embeddings for each image instead of time embeddings. The generated views are passed into a robust 3D reconstruction pipeline to create the 3D representation (Zip-NeRF or 3DGS)”

Bilawal Sidhu

92,792 görüntüleme • 2 yıl önce

NVidia just released PiD: super resolution in pixel space directly from model latents 🔎 4X resolution for any generated image, FAST! 🏎️💨 FLUX.1, 2 and Z-Image (Qwen Image coming) of course, i built a demo: generate 4K images with Z-Image
0:12

NVidia just released PiD: super resolution in pixel space directly from model latents 🔎 4X resolution for any generated image, FAST! 🏎️💨 FLUX.1, 2 and Z-Image (Qwen Image coming) of course, i built a demo: generate 4K images with Z-Image

apolinario (poli)

29,948 görüntüleme • 1 ay önce

Along with automatic skeletons, another thing I worked on recently was coordinated image loading. You can wrap any group of images with an &lt;ImageCoordinator&gt;, and all the images under it will fade in together rather than popcorning. Makes for a much more pleasant experience! 😀
0:12

Along with automatic skeletons, another thing I worked on recently was coordinated image loading. You can wrap any group of images with an &lt;ImageCoordinator&gt;, and all the images under it will fade in together rather than popcorning. Makes for a much more pleasant experience! 😀

Devon Govett

122,998 görüntüleme • 1 yıl önce

Yesterday someone suggested I create a fight between Alucard and Drolta Tzuentes using Kling AI 2.6. I usually work with Text to Video, but in this case I had to use Image to Video. The base image is a fusion of two images created in Niji and combined with GPT Image 1.5. Then I added the right prompt to give the fight speed and dynamism. The result isn’t perfect, but it’s pretty cool.
0:10

Yesterday someone suggested I create a fight between Alucard and Drolta Tzuentes using Kling AI 2.6. I usually work with Text to Video, but in this case I had to use Image to Video. The base image is a fusion of two images created in Niji and combined with GPT Image 1.5. Then I added the right prompt to give the fight speed and dynamism. The result isn’t perfect, but it’s pretty cool.

OscarAI

18,533 görüntüleme • 6 ay önce

I like my Laravel seeders fast. I use recycle() to dodge N+1 model creations. Another trick: when a model needs an image saved on disk, generating a unique one for each record is painfully slow. And honestly, images aren’t immediately essential. So I dispatch that work to a job. My local projects always run Horizon, so images get generated in the background while I keep coding. A local environment you can reset instantly with random data is a game-changer.
0:12

I like my Laravel seeders fast. I use recycle() to dodge N+1 model creations. Another trick: when a model needs an image saved on disk, generating a unique one for each record is painfully slow. And honestly, images aren’t immediately essential. So I dispatch that work to a job. My local projects always run Horizon, so images get generated in the background while I keep coding. A local environment you can reset instantly with random data is a game-changer.

Benjamin Crozat

10,488 görüntüleme • 11 ay önce

3. Modify image Use this prompt to modify the image you have chosen: "Modify image [1] with seed [1470033597]: add a parrot on her shoulder" Dall-E 3 identifies the image and makes the changes for you! Tips and tricks: - You can generate as many variations as you like. - It's also possible to remove an element from the image using the same method. - Sometimes the image isn't 100% identical, but still looks very similar.
0:10

3. Modify image Use this prompt to modify the image you have chosen: "Modify image [1] with seed [1470033597]: add a parrot on her shoulder" Dall-E 3 identifies the image and makes the changes for you! Tips and tricks: - You can generate as many variations as you like. - It's also possible to remove an element from the image using the same method. - Sometimes the image isn't 100% identical, but still looks very similar.

Paul Couvert

44,073 görüntüleme • 2 yıl önce

Trump to Husted: "He's running for a very important seat in a great state that I like a lot. It's always been good to me. And it's gonna be good to you, and it has been good to you. So good luck with that." (He's a senator from Ohio)
0:16

Trump to Husted: "He's running for a very important seat in a great state that I like a lot. It's always been good to me. And it's gonna be good to you, and it has been good to you. So good luck with that." (He's a senator from Ohio)

Aaron Rupar

50,213 görüntüleme • 4 ay önce

Impossible Nature. My goal here was to generate images with a low quality aesthetic that are slightly overexposed and then apply handheld camera motion to it with focal shifts, quick zoom in/out + tracking. I just wanted to see how far I could push realism from an alternate universe. Nano Banana Pro let's you construct animals out of practically anything, it's wild. What do you think? Does it look believable? Images generated with Google's Nano Banana Pro. Image to video using Kling AI 2.5 1 - A crab made entirely out of seashells captured on the beach at night.
0:10

Impossible Nature. My goal here was to generate images with a low quality aesthetic that are slightly overexposed and then apply handheld camera motion to it with focal shifts, quick zoom in/out + tracking. I just wanted to see how far I could push realism from an alternate universe. Nano Banana Pro let's you construct animals out of practically anything, it's wild. What do you think? Does it look believable? Images generated with Google's Nano Banana Pro. Image to video using Kling AI 2.5 1 - A crab made entirely out of seashells captured on the beach at night.

Travis Davids

16,750 görüntüleme • 7 ay önce

I created a prompt library for the new GPT image API. I vibe coded it this morning with Replit Agent. It allows me to save prompts as well as run multi-variant generations. The app supports batch uploads and has some other cool features: - Auth with Replit Auth - Object Storage for images - DB for users / prompts / sessions - Image compression for efficiency - Parallel calls to OpenAI for image gen
0:13

I created a prompt library for the new GPT image API. I vibe coded it this morning with Replit Agent. It allows me to save prompts as well as run multi-variant generations. The app supports batch uploads and has some other cool features: - Auth with Replit Auth - Object Storage for images - DB for users / prompts / sessions - Image compression for efficiency - Parallel calls to OpenAI for image gen

matt palmer

38,742 görüntüleme • 1 yıl önce

Hello artists, I have a problem! I'm looking for a way to fix this issue in Clip Studio Paint: As you can see in the video, my shadow layer is clipped to my base color layer, but it's still stretching the color to the edges of the base color. I've tried looking for an option or something to fix this, but haven't found anything. I've changed my brush settings, created a new brush, etc., but nothing works. I went back to Paint Tool SAI and saw that when I make a selection, it doesn't do this in Paint Tool SAI. So I tried it in Clip Studio Paint, and unfortunately, the problem persists. It spreads the color to the selection edges where Paint Tool SAI doesn't. If anyone has a solution, a better-programmed brush, an add-on, or anything that can fix this problem, I would be very grateful!
0:15

Hello artists, I have a problem! I'm looking for a way to fix this issue in Clip Studio Paint: As you can see in the video, my shadow layer is clipped to my base color layer, but it's still stretching the color to the edges of the base color. I've tried looking for an option or something to fix this, but haven't found anything. I've changed my brush settings, created a new brush, etc., but nothing works. I went back to Paint Tool SAI and saw that when I make a selection, it doesn't do this in Paint Tool SAI. So I tried it in Clip Studio Paint, and unfortunately, the problem persists. It spreads the color to the selection edges where Paint Tool SAI doesn't. If anyone has a solution, a better-programmed brush, an add-on, or anything that can fix this problem, I would be very grateful!

Aniu

51,316 görüntüleme • 7 ay önce

Midjourney sref + Sora 2 Pro is the sauce. With one Midjourney style image, you can give a specific style for your entire project. I created two different 12-second clips and edited them together. Some details aren’t fully consistent, like the iPod or AirPods because the clips were made separately from a single image (Character in a specific style). It could be fixed in post-production, but that would take more time, and this was more of an experimental test. It would be great to add the actual product image with the current one to maintain product consistency. I feel like if there were a way to add 2–4 images into this workflow, it could open up a lot more possibilities and consistency. With an API, it could be possible. Or let’s see what Veo 3.1 has to offer.
0:16

Midjourney sref + Sora 2 Pro is the sauce. With one Midjourney style image, you can give a specific style for your entire project. I created two different 12-second clips and edited them together. Some details aren’t fully consistent, like the iPod or AirPods because the clips were made separately from a single image (Character in a specific style). It could be fixed in post-production, but that would take more time, and this was more of an experimental test. It would be great to add the actual product image with the current one to maintain product consistency. I feel like if there were a way to add 2–4 images into this workflow, it could open up a lot more possibilities and consistency. With an API, it could be possible. Or let’s see what Veo 3.1 has to offer.

Allar Haltsonen

10,141 görüntüleme • 9 ay önce

Youtube demonetizing one-image videos??? Is it end for YouTube automation??? Nooo!!! Solution is 👇 SIMPLE: Create more images for a video. I hear you: "But it takes time..." Noooo! There is an AI tool — you just paste your script and it generates as many images as you want. Like + Retweet + Comment "tool" and I’ll send you the tool link.
0:11

Youtube demonetizing one-image videos??? Is it end for YouTube automation??? Nooo!!! Solution is 👇 SIMPLE: Create more images for a video. I hear you: "But it takes time..." Noooo! There is an AI tool — you just paste your script and it generates as many images as you want. Like + Retweet + Comment "tool" and I’ll send you the tool link.

aaron

37,816 görüntüleme • 11 ay 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.
0:12

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

52,801 görüntüleme • 1 yıl önce

Created a course that covers the fundamentals of starting an email business and creating an income online. I've been told it's better than most $997 courses! It's only FREE for the next 72 hours...will put a price on this soon. Follow, Like, RT, and reply "DM" to get access!
0:13

Created a course that covers the fundamentals of starting an email business and creating an income online. I've been told it's better than most $997 courses! It's only FREE for the next 72 hours...will put a price on this soon. Follow, Like, RT, and reply "DM" to get access!

Splash

86,325 görüntüleme • 3 yıl önce

This could be huge for AI storytelling. I've tried so many AI lipsync tools and have never been satisfied. This is my first try with Hedra Character-3 and it's blown me away. It was so easy. I uploaded a track from Suno and an image from Midjourney and it just worked.
0:10

This could be huge for AI storytelling. I've tried so many AI lipsync tools and have never been satisfied. This is my first try with Hedra Character-3 and it's blown me away. It was so easy. I uploaded a track from Suno and an image from Midjourney and it just worked.

TomLikesRobots🤖

42,723 görüntüleme • 1 yıl önce

GPT Image 2 + Seedance 2.0 Prompt Share Created on mitte.ai I didn't use a character sheet for this generation. I directly used the character images I created in Midjourney. Since the visual style transfers into the video surprisingly well, it's actually a really effective for style transfer. This time I also added a bit more detail to the Seedance prompt itself. You can definitely achieve similar results without storyboards too, they're not mandatory but I think they're one of the best ways to previsualize scenes, pacing and even camera angles before generation. Also, this storyboard prompt is still a bit long. I'm currently experimenting with more compact version of it too. You can check the prompts below.
0:14

GPT Image 2 + Seedance 2.0 Prompt Share Created on mitte.ai I didn't use a character sheet for this generation. I directly used the character images I created in Midjourney. Since the visual style transfers into the video surprisingly well, it's actually a really effective for style transfer. This time I also added a bit more detail to the Seedance prompt itself. You can definitely achieve similar results without storyboards too, they're not mandatory but I think they're one of the best ways to previsualize scenes, pacing and even camera angles before generation. Also, this storyboard prompt is still a bit long. I'm currently experimenting with more compact version of it too. You can check the prompts below.

Kōda

38,856 görüntüleme • 1 ay önce

This interface is a little "hidden" but you can access it easily. First, check that you have access to Google Search's AI functions. 1. Go to labs .google 2. Click on "Get started" in the Google Search category 3. Activate "SGE generative AI in Search" Next, write a prompt like I already showed you to draw an image in Google. Example: “draw a submarine in the sky” Once your images are generated, click on one then “Edit”.
0:13

This interface is a little "hidden" but you can access it easily. First, check that you have access to Google Search's AI functions. 1. Go to labs .google 2. Click on "Get started" in the Google Search category 3. Activate "SGE generative AI in Search" Next, write a prompt like I already showed you to draw an image in Google. Example: “draw a submarine in the sky” Once your images are generated, click on one then “Edit”.

Paul Couvert

127,585 görüntüleme • 2 yıl 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.
0:10

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 • 5 ay önce