Loading video...
Video Failed to Load
The (Processing) Level Required to Be Super We’ve had clients ask us whether our 3D/4DGS capture and processing pipeline holds up in high-end VFX cinematic shots. Last year we were fortunate to work on the Superman production with Framestore (capturing Superman’s parents for a hologram), with our work shown... show more
44,109 views • 10 months ago •via X (Twitter)
0 Comments
No comments available
Comments from the original post will appear here
![A Taiwanese reporter interviewed the PCES PC brand (the skirt Lisa wore during the Taiwan concert) about how they felt. One of the team members was so happy, they threw up! The brand showed great respect for Lisa, they won't sell the skirt but will put it in a frame for display. It's the brand's great pride, after all. “…. and the clothing brands they wore are still trending, just like the miniskirt Lisa wore from the Taiwanese brand PCES during the concert….” About 10 days before the concert started, we opened an email, and the subject line was written in large letters: BLACKPINK wants to borrow our clothing. At that time, we were stunned and frozen. We read the email twice and then showed it to our team. I was sitting on the floor working at the time, and they ran in and said, "Something huge has happened!" At that moment, the BP stylist asked us to provide more brand details and also check if we had other styles of clothing. This skirt is our newest, unreleased design….. About 15 minutes before the concert started, we were watching the time and then one of us exclaimed, "The concert is about to begin!" (Did you cry because you were so happy?) No, but one of our team members was so happy they threw up! 😂😂😂 It was a moment he had waited for for a long time; he was extremely excited. Q: (Did he really throw up?) - Yes, a little bit. We feel Lisa looked incredibly beautiful wearing it. Thank you so much, Lisa. Thank you to the BP team for paying attention to the details and choosing our brand—a genuine Taiwanese brand. Because this skirt hasn't been produced and put on sale yet, the one Lisa wore is the only sample piece in existence. Yesterday, the staff sent the item back. When we received it, we slowly opened it with care and took turns looking at it one by one. Q: (Can you tell us Lisa's waist size?) - Uh, the back of the skirt is adjustable in size. When we got the skirt back, it had been adjusted to be very small. We could see that Lisa adjusted it to a very small size. We think her waist might be 24 [inches], much smaller than the mannequin. (Fans must be asking if the skirt Lisa wore can be purchased/resold?) We probably won't sell it. It wouldn't be appropriate to sell it. It's something of sentimental value for our brand. We are going to buy a frame to keep it in and put it out on display. (It will become a precious treasure of the brand.) LLOUD #LISA](https://image.24vids.com/tw-1986616294574325815/amplify_video_thumb/1986616119558647808/img/I3Ss-T5XcVEWD18l.jpg)
![Everyone's sleeping on image-to-3D AI models. They can make your app look incredibly unique, with just a little effort. Here's how. This is my calorie tracker, built in a week with nothing but prompting. Just Claude Code + a couple APIs. The visuals are all AI-generated. I'll be sharing the full workflow + all the crazy technical stuff Claude and I did to make this work, so nobody has to struggle through it like me. Deep dive coming soon! Till then, this is the high-level idea: 1. Get a clean image of the food (or whatever your asset is) - In my app, the user describes foods via text, or attaches images (or both) - If text, an LLM extracts the food description and formats it into a specific prompt I tuned for this design, and we generate an image using Z-Image Turbo through fal - If image, we do the same thing but with FLUX.2 [dev] to edit the user image into our reference design - Originally, both used Google Nano Banana, but switching to open models cut costs and latency a ton 2. Gaussian splatting (2D image → 3D model) - I tried various 2D-to-3D options on fal and ended up with TripoSplat as my preferred balance of speed, cost, latency; this turns an image into a 3D model that looks super high quality (link below) - The app displays the 2D image while our backend generates the 3D splat - We "groom" the splat to reduce size and load time by culling low-opacity/scale points 3. Render efficiently on device Originally, it looked great but ran at 10 FPS. Getting to 120 FPS was a crazy journey. TL;DR: - SwiftUI had to go; it forced us to render each asset in independent MTKViews, which wasn't workable - Instead, we composite every dish into one full-bleed CAMetalLayer using MetalSplatter (link below) - We had to make some optimizations within MetalSplatter's code too, to reduce the overhead of sorting points per render Then I added some finishing touches like the subtle rotation and parallax as they move around. I think it turned out pretty cool :) Overall, this took some effort, but we still got it done in less than a day. Hopefully your agent can follow in the footsteps of mine and do it much faster. Keep an eye out for the bigger writeup, which'll give your agent everything it needs. If you have any questions, drop em below!](https://image.24vids.com/tw-2069277649655243140/amplify_video_thumb/2069276621698465792/img/7_y1PzRYfzBgLDJ6.jpg)
