Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

Here's an actual design engineering tip: Safari sucks at updating multiple WebGL video textures in the same frame, especially during interactions or animations on initial page load. To keep things buttery you'll wanna pause texture uploads during motion e.g: isMoving ? holdLastFrame() : gl.texImage2D(video)

joshpuckett's profile picture

joshpuckett

41,266 subscribers

36,713 Aufrufe • vor 15 Tagen •via X (Twitter)

Wissenschaft & TechnologieBildungComedy
LIVE

Anya Rossi• Live Now

Private livecam show

0 Kommentare

Keine Kommentare verfügbar

Kommentare vom Original-Post werden hier angezeigt

Ähnliche Videos

Here's a tip I don't see enough: make your animations interruptible. You should be able to immediately trigger close events without waiting for the animation to complete. Motion makes this super easy. This helps your interfaces feel durable, snappy, and well considered.
0:11

Here's a tip I don't see enough: make your animations interruptible. You should be able to immediately trigger close events without waiting for the animation to complete. Motion makes this super easy. This helps your interfaces feel durable, snappy, and well considered.

joshpuckett

101,682 Aufrufe • vor 8 Monaten

Rerender A Video: Zero-Shot Text-Guided Video-to-Video Translation paper page: Large text-to-image diffusion models have exhibited impressive proficiency in generating high-quality images. However, when applying these models to video domain, ensuring temporal consistency across video frames remains a formidable challenge. This paper proposes a novel zero-shot text-guided video-to-video translation framework to adapt image models to videos. The framework includes two parts: key frame translation and full video translation. The first part uses an adapted diffusion model to generate key frames, with hierarchical cross-frame constraints applied to enforce coherence in shapes, textures and colors. The second part propagates the key frames to other frames with temporal-aware patch matching and frame blending. Our framework achieves global style and local texture temporal consistency at a low cost (without re-training or optimization). The adaptation is compatible with existing image diffusion techniques, allowing our framework to take advantage of them, such as customizing a specific subject with LoRA, and introducing extra spatial guidance with ControlNet. Extensive experimental results demonstrate the effectiveness of our proposed framework over existing methods in rendering high-quality and temporally-coherent videos.
0:17

Rerender A Video: Zero-Shot Text-Guided Video-to-Video Translation paper page: Large text-to-image diffusion models have exhibited impressive proficiency in generating high-quality images. However, when applying these models to video domain, ensuring temporal consistency across video frames remains a formidable challenge. This paper proposes a novel zero-shot text-guided video-to-video translation framework to adapt image models to videos. The framework includes two parts: key frame translation and full video translation. The first part uses an adapted diffusion model to generate key frames, with hierarchical cross-frame constraints applied to enforce coherence in shapes, textures and colors. The second part propagates the key frames to other frames with temporal-aware patch matching and frame blending. Our framework achieves global style and local texture temporal consistency at a low cost (without re-training or optimization). The adaptation is compatible with existing image diffusion techniques, allowing our framework to take advantage of them, such as customizing a specific subject with LoRA, and introducing extra spatial guidance with ControlNet. Extensive experimental results demonstrate the effectiveness of our proposed framework over existing methods in rendering high-quality and temporally-coherent videos.

AK

375,160 Aufrufe • vor 3 Jahren

The cool thing about animating in code is it's pretty trivial to do quick variants like this x-ray inversion.* 15 hours left on this! *not sponsored by, affiliated with, or anywhere near the quality of Shopify's masterful design site
0:10

The cool thing about animating in code is it's pretty trivial to do quick variants like this x-ray inversion.* 15 hours left on this! *not sponsored by, affiliated with, or anywhere near the quality of Shopify's masterful design site

joshpuckett

34,975 Aufrufe • vor 5 Monaten

Good software is durable. Here's an quick example of what that means, at an interface level. In X, you can choose a custom emoji reaction. The popover looks like what I expect, but it does not work like I expect. I want to use the ‼ emoji, so the first thing I try is obviously '!'. No results. The sum of a million little papercuts like this means X feels like broken software, not well designed or made. Which is a shame, because it's a brilliant product and network.
0:13

Good software is durable. Here's an quick example of what that means, at an interface level. In X, you can choose a custom emoji reaction. The popover looks like what I expect, but it does not work like I expect. I want to use the ‼ emoji, so the first thing I try is obviously '!'. No results. The sum of a million little papercuts like this means X feels like broken software, not well designed or made. Which is a shame, because it's a brilliant product and network.

joshpuckett

55,420 Aufrufe • vor 8 Monaten

A few people have asked how I make these animations. The answer is it's just code and a screen recording. I’ll make little jigs like this interactive timeline to dial things in, but for me I find this is still the fastest way to create and refine exactly what I want.
0:10

A few people have asked how I make these animations. The answer is it's just code and a screen recording. I’ll make little jigs like this interactive timeline to dial things in, but for me I find this is still the fastest way to create and refine exactly what I want.

joshpuckett

53,125 Aufrufe • vor 2 Monaten

this AI agent runs multiple merch brands on tiktok and instagram, from slack...on autopilot. here's how solopreneurs can run a viral merch business without touching design or video: - scans tiktok and instagram for what's actually going viral - turns each trend into a short-form video script - generates an AI video that matches each trend - posts to every platform at peak times - designs matching merch, lists it live, links in the pinned comment - replies to every comment, kills duds, scales winners every step from a viral trend to a sold shirt is automated reply "MERCH" + RT and i'll send you the full setup so you can run this too
0:11

this AI agent runs multiple merch brands on tiktok and instagram, from slack...on autopilot. here's how solopreneurs can run a viral merch business without touching design or video: - scans tiktok and instagram for what's actually going viral - turns each trend into a short-form video script - generates an AI video that matches each trend - posts to every platform at peak times - designs matching merch, lists it live, links in the pinned comment - replies to every comment, kills duds, scales winners every step from a viral trend to a sold shirt is automated reply "MERCH" + RT and i'll send you the full setup so you can run this too

Chris

13,471 Aufrufe • vor 4 Monaten

The combat you saw in the Content creators demo was just a really small sneak peek… Let me show you more about the real combat of AIKODE… In this video there is A LOT of things going on, in fact, the outfit changed 4 times in this video! In AIKODE you can change between 3 Outfits during GAMEPLAY (you can choose what 3 outfits will you use in the menu between 200 different options). The Outfit system also changes your weapon, animations, hairstyle and abilities, (You can customize all of this on the menu for every outfit slot). While the abilities are “charging” you can keep changing between outfits to use the abilities that are already charged... and at the same time you can keep doing combos without any type of delay and smoothly change between your weapons. In this video you can see just some of the abilities, outfits, and stances/weapons. #AIKODE #UE5
0:17

The combat you saw in the Content creators demo was just a really small sneak peek… Let me show you more about the real combat of AIKODE… In this video there is A LOT of things going on, in fact, the outfit changed 4 times in this video! In AIKODE you can change between 3 Outfits during GAMEPLAY (you can choose what 3 outfits will you use in the menu between 200 different options). The Outfit system also changes your weapon, animations, hairstyle and abilities, (You can customize all of this on the menu for every outfit slot). While the abilities are “charging” you can keep changing between outfits to use the abilities that are already charged... and at the same time you can keep doing combos without any type of delay and smoothly change between your weapons. In this video you can see just some of the abilities, outfits, and stances/weapons. #AIKODE #UE5

AIKODE

62,062 Aufrufe • vor 2 Jahren

One of my favorite details for Interface Craft is one of the least used: the invoice tool. Customers can enter the unique tax or business info they need, see what they paid in localized currency along with any discounts, and download a PDF. Thankfully Stripe makes this all fairly easy. It's something only a fraction of users see, but I think great design is sometimes about facilitating all of the edge cases to keep accounting and tax departments happy worldwide!
0:10

One of my favorite details for Interface Craft is one of the least used: the invoice tool. Customers can enter the unique tax or business info they need, see what they paid in localized currency along with any discounts, and download a PDF. Thankfully Stripe makes this all fairly easy. It's something only a fraction of users see, but I think great design is sometimes about facilitating all of the edge cases to keep accounting and tax departments happy worldwide!

joshpuckett

67,651 Aufrufe • vor 2 Monaten

I'm hosting a private tour of the Matisse exhibit at SFMOMA on Tuesday, June 23rd. I hope it'll be an inspiring way to kick off Config week in San Francisco! Spots are limited; apply here:
0:14

I'm hosting a private tour of the Matisse exhibit at SFMOMA on Tuesday, June 23rd. I hope it'll be an inspiring way to kick off Config week in San Francisco! Spots are limited; apply here:

joshpuckett

45,128 Aufrufe • vor 3 Monaten

🚨 so let me get this straight.. Biden had the intelligence files on Tren de Aragua.. had the Foreign Terrorist Organization designation authority.. had US Southern Command assets in position.. and he waited.. Trump takes office.. designates TDA a Foreign Terrorist Organization.. authorizes strikes.. kills 11 members at sea in 2025.. then kills the leader himself in 2026.. Hector Rusthenford Guerrero Flores.. Niño Guerrero.. head of one of the most violent criminal organizations on earth.. operating across 40+ countries.. untouched for years.. gone.. confirmed killed.. with unclassified overhead video of the building exploding released the same night.. and here's the part nobody's connecting.. Trump announced it during the opening USA World Cup match.. while 100 million Americans were watching the game.. with the video.. publicly.. on Truth Social.. you don't release overhead strike footage during the World Cup opener by accident.. every cartel and criminal network watching the same screen got the same message at the same time.. this is the same playbook every single time.. MOAB in 2017.. embassy move in 2018.. Soleimani in 2020.. the file was always there.. Trump just treats every deferred file as an immediate action item.. the "too dangerous to touch" era just ended on the night of the World Cup opener. I'll keep you updated as this unfolds, turn on notifications this is EXTREMELY important.
0:10

🚨 so let me get this straight.. Biden had the intelligence files on Tren de Aragua.. had the Foreign Terrorist Organization designation authority.. had US Southern Command assets in position.. and he waited.. Trump takes office.. designates TDA a Foreign Terrorist Organization.. authorizes strikes.. kills 11 members at sea in 2025.. then kills the leader himself in 2026.. Hector Rusthenford Guerrero Flores.. Niño Guerrero.. head of one of the most violent criminal organizations on earth.. operating across 40+ countries.. untouched for years.. gone.. confirmed killed.. with unclassified overhead video of the building exploding released the same night.. and here's the part nobody's connecting.. Trump announced it during the opening USA World Cup match.. while 100 million Americans were watching the game.. with the video.. publicly.. on Truth Social.. you don't release overhead strike footage during the World Cup opener by accident.. every cartel and criminal network watching the same screen got the same message at the same time.. this is the same playbook every single time.. MOAB in 2017.. embassy move in 2018.. Soleimani in 2020.. the file was always there.. Trump just treats every deferred file as an immediate action item.. the "too dangerous to touch" era just ended on the night of the World Cup opener. I'll keep you updated as this unfolds, turn on notifications this is EXTREMELY important.

🇺🇸 Edward T. Winslow

83,258 Aufrufe • vor 2 Monaten

The video smooth zoom on the Samsung Galaxy S26 Ultra is still the closest thing to a professional camcorder experience in the smartphone industry today. In fact, it’s even easier to control than the iPhone. On many other phones, video zooming requires constant finger movement and very precise control. The zoom speed can easily become inconsistent, suddenly speeding up or slowing down. Samsung works differently. You simply hold your finger at a certain position, and the phone continues zooming at a constant speed. The entire process feels extremely stable and linear. It genuinely resembles the powered zoom control of a professional video camera. This logic is fundamentally related to Samsung’s AI slow motion technology. They share the same core foundation: real time control over motion trajectories, speed transitions, and frame interpolation. What you’re seeing here was shot in very windy conditions using Samsung’s Pro Video mode, continuously zooming from 5x to 25x. Aside from some slight stutter during optical lens switching points, the continuous zoom transition within digital zoom ranges is arguably the closest thing to a professional camera currently available on a smartphone. So if the future Samsung Galaxy S27 Ultra really removes the 3x telephoto camera, it could actually improve the video zoom experience further. Fewer optical switching points would theoretically reduce transition jumps and stutters, making the entire zoom range feel even more natural and continuous.
0:13

The video smooth zoom on the Samsung Galaxy S26 Ultra is still the closest thing to a professional camcorder experience in the smartphone industry today. In fact, it’s even easier to control than the iPhone. On many other phones, video zooming requires constant finger movement and very precise control. The zoom speed can easily become inconsistent, suddenly speeding up or slowing down. Samsung works differently. You simply hold your finger at a certain position, and the phone continues zooming at a constant speed. The entire process feels extremely stable and linear. It genuinely resembles the powered zoom control of a professional video camera. This logic is fundamentally related to Samsung’s AI slow motion technology. They share the same core foundation: real time control over motion trajectories, speed transitions, and frame interpolation. What you’re seeing here was shot in very windy conditions using Samsung’s Pro Video mode, continuously zooming from 5x to 25x. Aside from some slight stutter during optical lens switching points, the continuous zoom transition within digital zoom ranges is arguably the closest thing to a professional camera currently available on a smartphone. So if the future Samsung Galaxy S27 Ultra really removes the 3x telephoto camera, it could actually improve the video zoom experience further. Fewer optical switching points would theoretically reduce transition jumps and stutters, making the entire zoom range feel even more natural and continuous.

Ice Universe

27,250 Aufrufe • vor 4 Monaten

This is the easiest way to make $10k/month with organic affiliate and AI Arcads launched an ai ugc studio that lets you build an entire army of hyper-real AI actors Then you turn any static image into a high-quality video showcasing any product go to TikTok and make an account + warm it up using arcads you can run an entirely AI UGC account using the same character over and over, making it seem like an authentic TT page Mix the content up with slideshows and videos with the same character Here's the AI stack gameplan: - Claude to help you write scripts - Arcads to generate an image of an AI girlie that fits your product demographic Scroll tiktok and save + download every video / slideshow you see made by clippers promoting a product (there's literally loads) Your going to find an offer on whop for making money online or spirituality and target it towards girls feed all these videos you scraped into a custom google gemini gem trained to deconstruct hooks / angles for you for easy hook inspiration + ideas Deconstruct the hooks, put them into Claude and ask it to give you hooks for the same style of video put for your products your promoting For the videos do caption and reaction + showcase formats Generate the reactions using the character you made in arc ads then manually record the showcasing of the product or proof of the product working Also for caption generate a 8-10 second video you can put text over Include your CTA in the video for reaction style and captions for caption style Plus generate images with the same character and make slideshows directed to your product Now rinse and repeat this make multiple accounts with multiple different avatars and print
0:14

This is the easiest way to make $10k/month with organic affiliate and AI Arcads launched an ai ugc studio that lets you build an entire army of hyper-real AI actors Then you turn any static image into a high-quality video showcasing any product go to TikTok and make an account + warm it up using arcads you can run an entirely AI UGC account using the same character over and over, making it seem like an authentic TT page Mix the content up with slideshows and videos with the same character Here's the AI stack gameplan: - Claude to help you write scripts - Arcads to generate an image of an AI girlie that fits your product demographic Scroll tiktok and save + download every video / slideshow you see made by clippers promoting a product (there's literally loads) Your going to find an offer on whop for making money online or spirituality and target it towards girls feed all these videos you scraped into a custom google gemini gem trained to deconstruct hooks / angles for you for easy hook inspiration + ideas Deconstruct the hooks, put them into Claude and ask it to give you hooks for the same style of video put for your products your promoting For the videos do caption and reaction + showcase formats Generate the reactions using the character you made in arc ads then manually record the showcasing of the product or proof of the product working Also for caption generate a 8-10 second video you can put text over Include your CTA in the video for reaction style and captions for caption style Plus generate images with the same character and make slideshows directed to your product Now rinse and repeat this make multiple accounts with multiple different avatars and print

Pounds

32,407 Aufrufe • vor 7 Monaten

Means & Methods is now live on A collection of practical techniques to achieve excellence in interface design, covering 100+ topics across 11 chapters, with plenty of interactive examples and code. New members welcome; enjoy!
0:18

Means & Methods is now live on A collection of practical techniques to achieve excellence in interface design, covering 100+ topics across 11 chapters, with plenty of interactive examples and code. New members welcome; enjoy!

joshpuckett

703,735 Aufrufe • vor 2 Monaten

I built a clone of the Yeezy store with Next.js. This was a fun challenge — the site has some smooth animations and feels very fast. But it was bothering me that I couldn't use the browser back button. Can we do better? So I rebuilt the site with v0 and Motion. Here's how it works: 1. When you click on a product, Motion is able to animate the original position of the product in the grid, to the zoomed in product detail page. 2. During this transition, we also shallow update the URL with the `/p/slug` route for the page. 3. If you press the back button in the navbar, or use the browser back button, or press escape — all options will take you back to the main product listing page. 4. If you reload the page while looking at a product, or someone sends you a link to a specific product, it still works! This is the best parts of a SPA and MPA mixed together. In the future, I can make this even better with View Transitions (I wasn't able to get the product animation just right, but if you can I'd love to see it!). I also took some creative liberties from the original design. The whole 1/2/3 size thing, where you needed to click the "?" to see SM/MD/LG was strange, so I just went directly to those sizes. Similarly, I prefered the more traditional style sheet/modal with the background color change, versus the full screen takeover. If you wanted to actually hook this up to Shopify now, you can swap the cart implementation with Next.js Commerce, which has all the APIs you need + optimistic writes 🔥 Should I make a video walking through the code?
0:16

I built a clone of the Yeezy store with Next.js. This was a fun challenge — the site has some smooth animations and feels very fast. But it was bothering me that I couldn't use the browser back button. Can we do better? So I rebuilt the site with v0 and Motion. Here's how it works: 1. When you click on a product, Motion is able to animate the original position of the product in the grid, to the zoomed in product detail page. 2. During this transition, we also shallow update the URL with the `/p/slug` route for the page. 3. If you press the back button in the navbar, or use the browser back button, or press escape — all options will take you back to the main product listing page. 4. If you reload the page while looking at a product, or someone sends you a link to a specific product, it still works! This is the best parts of a SPA and MPA mixed together. In the future, I can make this even better with View Transitions (I wasn't able to get the product animation just right, but if you can I'd love to see it!). I also took some creative liberties from the original design. The whole 1/2/3 size thing, where you needed to click the "?" to see SM/MD/LG was strange, so I just went directly to those sizes. Similarly, I prefered the more traditional style sheet/modal with the background color change, versus the full screen takeover. If you wanted to actually hook this up to Shopify now, you can swap the cart implementation with Next.js Commerce, which has all the APIs you need + optimistic writes 🔥 Should I make a video walking through the code?

Lee Robinson

148,937 Aufrufe • vor 1 Jahr

It’s just a video player on localhost:8080. That’s what I keep telling myself. Minimax H3 on Fish Creative Prompt Fixed-camera desktop screen recording of a real Windows 11 Chrome window filling the entire frame. Do not restyle the browser as a mockup, poster, or floating card. Keep the exact chrome from the reference still for every frame. Browser frame (locked, zero motion): - Tab title: Local Video Player - Address bar URL: localhost:8080 - Bookmarks bar left to right: Apps, Work, Personal, Google, GitHub, Docs, YouTube, then Other bookmarks on the right - Chrome controls: back, forward, refresh, extensions, profile avatar “M”, puzzle piece, shield, three-dot menu - Windows 11 taskbar along the bottom: search “Type here to search”, File Explorer, Google Chrome, Visual Studio Code, system tray, time 10:30 AM, date 5/20/2024 Page layout (labels, colors, and positions never change): - Left: large HTML5 video canvas - Under the canvas: play/pause triangle, timecode starting 00:00 / 00:10, purple seek bar, speaker, purple volume slider, fullscreen icon - Right sidebar on near-black: - “Video Player” with purple clapperboard icon - Controls: Space Play / Pause; ← → Seek ±5 seconds; ↑ ↓ Volume ±10%; F Toggle Fullscreen; M Mute / Unmute - Playlist: purple play arrow, “Sample Video”, 00:10 - Playback Rate: 0.5x, 1x filled purple and selected, 1.5x, 2x Character identity (must stay the same person in every pose): Photoreal young woman, long wavy blonde hair, gray school-style blazer over a white collared shirt, gray pleated mini skirt, black knee-high socks. Same face, same proportions, same outfit. Location stays the green grassy hill under a pale overcast sky. No costume change, no second character, no extra props. Motion rules: - Camera never moves. Browser chrome, sidebar, bookmarks, taskbar, and all UI text stay perfectly still. - Only the video canvas animates. Treat the canvas as a short fashion / pose test clip of the same girl on the hill. - Playback runs continuously at 1x. Do not pause the player. Do not flash a big pause icon. Do not cut away from the hill. - Timecode and the purple progress thumb advance smoothly from 00:00 to 00:10. Pose sequence inside the canvas (hold each pose ~1.5–2 seconds, then transition with natural body motion, hair, and cloth, not a hard cut): 1. Start pose (match the reference still): leaning forward, both hands on her knees, looking into camera with a pout. 2. Rise and stand tall: she straightens up, hands leave her knees, stands facing camera with arms relaxed at her sides, chin slightly lifted, wind moving her hair. 3. Hands-on-hips: she plants both hands on her hips, weight on one leg, slight hip cock, blazer and skirt settling, confident look into camera. 4. Turn / three-quarter: she twists her torso to her right, looks back over her shoulder toward camera, hair swinging, one hand lightly holding the hem of the skirt so it does not lift unnaturally. 5. Crouch / ready: she drops into a low athletic crouch, fingertips brushing the grass, still looking up at camera, same pout softening into a small smirk. 6. End pose: she stands again, steps one foot forward, both hands clasp in front of her thighs, then eases back toward the original lean-with-hands-on-knees so the last frame rhymes with the first still. Transitions must look like the same live-action take: weight shifts, knee bend, hair lag, grass and sky stay consistent. No teleport, no outfit swap, no extra readable text burned into the video. Audio: quiet outdoor hill ambience only. No music, no voiceover, no UI click spam. Style: photoreal live-action screen capture of a local demo page running in Chrome. Not animation, not a trailer, not a poster. Duration: about 10 seconds. Aspect: 16:9 desktop.
0:10

It’s just a video player on localhost:8080. That’s what I keep telling myself. Minimax H3 on Fish Creative Prompt Fixed-camera desktop screen recording of a real Windows 11 Chrome window filling the entire frame. Do not restyle the browser as a mockup, poster, or floating card. Keep the exact chrome from the reference still for every frame. Browser frame (locked, zero motion): - Tab title: Local Video Player - Address bar URL: localhost:8080 - Bookmarks bar left to right: Apps, Work, Personal, Google, GitHub, Docs, YouTube, then Other bookmarks on the right - Chrome controls: back, forward, refresh, extensions, profile avatar “M”, puzzle piece, shield, three-dot menu - Windows 11 taskbar along the bottom: search “Type here to search”, File Explorer, Google Chrome, Visual Studio Code, system tray, time 10:30 AM, date 5/20/2024 Page layout (labels, colors, and positions never change): - Left: large HTML5 video canvas - Under the canvas: play/pause triangle, timecode starting 00:00 / 00:10, purple seek bar, speaker, purple volume slider, fullscreen icon - Right sidebar on near-black: - “Video Player” with purple clapperboard icon - Controls: Space Play / Pause; ← → Seek ±5 seconds; ↑ ↓ Volume ±10%; F Toggle Fullscreen; M Mute / Unmute - Playlist: purple play arrow, “Sample Video”, 00:10 - Playback Rate: 0.5x, 1x filled purple and selected, 1.5x, 2x Character identity (must stay the same person in every pose): Photoreal young woman, long wavy blonde hair, gray school-style blazer over a white collared shirt, gray pleated mini skirt, black knee-high socks. Same face, same proportions, same outfit. Location stays the green grassy hill under a pale overcast sky. No costume change, no second character, no extra props. Motion rules: - Camera never moves. Browser chrome, sidebar, bookmarks, taskbar, and all UI text stay perfectly still. - Only the video canvas animates. Treat the canvas as a short fashion / pose test clip of the same girl on the hill. - Playback runs continuously at 1x. Do not pause the player. Do not flash a big pause icon. Do not cut away from the hill. - Timecode and the purple progress thumb advance smoothly from 00:00 to 00:10. Pose sequence inside the canvas (hold each pose ~1.5–2 seconds, then transition with natural body motion, hair, and cloth, not a hard cut): 1. Start pose (match the reference still): leaning forward, both hands on her knees, looking into camera with a pout. 2. Rise and stand tall: she straightens up, hands leave her knees, stands facing camera with arms relaxed at her sides, chin slightly lifted, wind moving her hair. 3. Hands-on-hips: she plants both hands on her hips, weight on one leg, slight hip cock, blazer and skirt settling, confident look into camera. 4. Turn / three-quarter: she twists her torso to her right, looks back over her shoulder toward camera, hair swinging, one hand lightly holding the hem of the skirt so it does not lift unnaturally. 5. Crouch / ready: she drops into a low athletic crouch, fingertips brushing the grass, still looking up at camera, same pout softening into a small smirk. 6. End pose: she stands again, steps one foot forward, both hands clasp in front of her thighs, then eases back toward the original lean-with-hands-on-knees so the last frame rhymes with the first still. Transitions must look like the same live-action take: weight shifts, knee bend, hair lag, grass and sky stay consistent. No teleport, no outfit swap, no extra readable text burned into the video. Audio: quiet outdoor hill ambience only. No music, no voiceover, no UI click spam. Style: photoreal live-action screen capture of a local demo page running in Chrome. Not animation, not a trailer, not a poster. Duration: about 10 seconds. Aspect: 16:9 desktop.

Sharon Riley

23,360 Aufrufe • vor 5 Tagen

The easiest way to make something good is to quickly make a lot of bad things first. It's all about iteration...
0:11

The easiest way to make something good is to quickly make a lot of bad things first. It's all about iteration...

joshpuckett

29,065 Aufrufe • vor 2 Monaten

I've seen some conservatives try to start an Obama Murder Conspiracy theory because, well, they think they are so smart. Obama did go golfing and had what appears to be bandaids on his fingers after his chef's death. Here's where your story falls apart: If you even bothered to look at past pictures of Obama golfing, these are sports bandages, that he actually happens to wear in the exact same spots during golfing outings as far back as 7 years ago. Golfers wear these often in order to prevent blisters from swinging. They are an alternative to wearing sweaty gloves. As Evidence, you can see the video below showing past times Obama wore the same 'bandages'. It also appears that Tiger Woods may be a murderer. Unless he's a serial killer that keeps getting hurt in the same spot whenever he murders someone, your conspiracy theories have NO MERIT!
0:14

I've seen some conservatives try to start an Obama Murder Conspiracy theory because, well, they think they are so smart. Obama did go golfing and had what appears to be bandaids on his fingers after his chef's death. Here's where your story falls apart: If you even bothered to look at past pictures of Obama golfing, these are sports bandages, that he actually happens to wear in the exact same spots during golfing outings as far back as 7 years ago. Golfers wear these often in order to prevent blisters from swinging. They are an alternative to wearing sweaty gloves. As Evidence, you can see the video below showing past times Obama wore the same 'bandages'. It also appears that Tiger Woods may be a murderer. Unless he's a serial killer that keeps getting hurt in the same spot whenever he murders someone, your conspiracy theories have NO MERIT!

Brian Krassenstein

1,992,442 Aufrufe • vor 3 Jahren

are we still allowed to have fun on the internet in 2026? 🎶
0:10

are we still allowed to have fun on the internet in 2026? 🎶

joshpuckett

382,112 Aufrufe • vor 1 Monat

You can't 3D reconstruct glass from images... ...WRONG! Thanks for video diffusion, now just about anything is possible! Introducing...Diffusion Knows Transparency (DKT) Transparent and reflective objects usually break robot vision and photogrammetry pipelines because they don't follow the "solid object" rules standard cameras expect. DKT is a new AI model that repurposes the "internal physics engine" found in video generation models to solve this problem. Researchers took a massive video diffusion model (WAN) and fine-tuned it using a custom-built synthetic dataset to turn it into a high-precision depth sensor. To train the AI, they built the first massive synthetic video library of transparent objects, 1.32 million frames of perfectly labeled glass and metal objects in motion. Without ever seeing a "real" labeled video of glass during training, the model (DKT) outperformed all previous specialized systems on real-world benchmarks (ClearPose, DREDS). They created a "lightweight" 1.3B parameter version that runs fast enough (0.17s per frame) to be used on actual robot hardware. Two reasons I find this project important: 1. It further proves that synthetic data will be essential for training the next generation vision models. 2. In real-world robotic tests, using DKT's depth maps nearly doubled the success rate of robot arms trying to pick up objects on tricky reflective or translucent surfaces. At home robots will need to interact with these types of objects on a daily basis. Check out the project page here: Code is LIVE! #Computervision #Robotics #AI
0:14

You can't 3D reconstruct glass from images... ...WRONG! Thanks for video diffusion, now just about anything is possible! Introducing...Diffusion Knows Transparency (DKT) Transparent and reflective objects usually break robot vision and photogrammetry pipelines because they don't follow the "solid object" rules standard cameras expect. DKT is a new AI model that repurposes the "internal physics engine" found in video generation models to solve this problem. Researchers took a massive video diffusion model (WAN) and fine-tuned it using a custom-built synthetic dataset to turn it into a high-precision depth sensor. To train the AI, they built the first massive synthetic video library of transparent objects, 1.32 million frames of perfectly labeled glass and metal objects in motion. Without ever seeing a "real" labeled video of glass during training, the model (DKT) outperformed all previous specialized systems on real-world benchmarks (ClearPose, DREDS). They created a "lightweight" 1.3B parameter version that runs fast enough (0.17s per frame) to be used on actual robot hardware. Two reasons I find this project important: 1. It further proves that synthetic data will be essential for training the next generation vision models. 2. In real-world robotic tests, using DKT's depth maps nearly doubled the success rate of robot arms trying to pick up objects on tricky reflective or translucent surfaces. At home robots will need to interact with these types of objects on a daily basis. Check out the project page here: Code is LIVE! #Computervision #Robotics #AI

Jonathan Stephens

17,712 Aufrufe • vor 8 Monaten