ap 🦖

ColorVideoVDP-ML was awarded second place in the full reference category of ICME Generalizable HDR & SDR Video Quality. The new metric models both the low-level (ColorVideoVDP) and high-level (ML) human vision. Now the code is available at github.com/gfxdisp/ColorV…

Who did this?

🆕The Workshop on Computer Vision for Mixed Reality @CVPR 2025: 📜Papers live here: cv4mr.github.io 🎁Speakers: Richard Newcombe @rapideRobot Anjul Patney @killerooo Rana Hanocka @RanaHanocka Laura Leal-Taixe Margarita Grinvald 📅Jun 11, 8 AM-12:15PM 📍 Room 109 ✌️

🚀 #HPG2025 schedule is live! 🔗 highperformancegraphics.org/2025/program/s… Dive into 3 days of real-time graphics: Keynotes (AMD, Roblox, NVIDIA) Paper & poster sessions Hot3D talks Student competition Harbour Boat Tour & Banquet Register now!!

I'm proud of our team's contributions to bringing DLSS tech to Switch 2! It was a lot of work to bring AI accelerated graphics to a new platform, but collaboration between multiple engineering and research teams made it a breeze. blogs.nvidia.com/blog/nintendo-…

@MittringMartin You can also unzip .pptx files to extract all raw assets.

If you want to export an image from Google Slides: Download slides as .pptx. Open in PowerPoint. Right mouse click on image. "Save As Picture...". Click on triangle next to "Save" to choose "Save Original Picture". Otherwise you get a low res version. "Save to Keep" is bad too.

RTX On 👍

Modern tech company logos if they existed in the 1980s

If anything, DeepSeek will unlock the plateauing that LLM advances had experienced over the last few months. That's a better news for the industry than the "AI is slowing down and there's no more data to learn from." Personal opinion.

Reflex 2 Frame Warp has been announced. Check it out! youtube.com/watch?v=zpDxo2…

@Peter_shirley @hdrsci might have many/all of them.

I am in search of all the classic HDR images people use in tone mapping comparisons. Anyone got a stash or a pointer. (yes I know I have had them all before... let's not discuss :D )

A little late, but just as excited. 20 years since #halflife2 landed and #AlyxVance made her way into my heart. The first time I got to play true to my ethnicity of a Black/Asian mix. I love her always. Thank you for the years of love. It will always mean the world.

@jbhuang0604 Feeling so disillusioned by the replies. Is a word worth a thousand pictures now?

Why is self-supervision in vision still not working? 🤔 When pretraining a transformer on TEXT-only data by predicting the next tokens, we see clear improvement trends as we scale the model, data, and computing. But after trying to pretrain a transformer on IMAGES-only data with contrastive learning (e.g., SimCLR), masked prediction (e.g., MAE), and self-distillation (e.g., Dino), at the end of the day, the most successful vision encoders still have to rely on vision-language paired data (e.g., CLIP. SigLIP). Or have people not yet tried connecting MAE/Dino with LLMs? Why is that? Thoughts?

Competition is good.

@omershapira Would you program it in a box? Would you do it with a fox?

I will never program a gate array, indoors or outdoors – not to mention in a field. disgusting

@niallw_cs Congratulations @niallw_cs !

@RuthRosenholtz "Attention is all you need"

Husband and I are thinking of getting a parrot-like bird. We’ll teach it to say common phrases in our household. “I think an ideal observer predicts that.” “Turn off auto scaling.” “Occam’s Razor.” “Did you use an eye tracker or just ask the subject to fixate?”

@Peter_shirley You forgot the temporal consistency loss function!

TAA and regular sampling artifacts (real video)

We finally get to share our neural materials research. We created custom highly-realistic reference materials, we invented new neural models, and we built new graphics systems. End result is these neural materials run in a real-time path tracer. research.nvidia.com/labs/rtr/neura…

“As a PC gamer I’ve been taking AI for granted for way too long — that stops now. DLSS, I love you” <- We love you too! :) tomsguide.com/gaming/as-a-pc…