ML 101:
data = food
docker = environment
code = home
compute = time
model = child
training = sleep
inference = talking
In brightest day,
In darkest night
I'll embrace my fears,
I'll do what's right.
I see the way,
I choose this life
Inspired by
The Sun's starlight.
Ilya Zakharkin retweeted
Found some time to add scene collisions to my Vision Pro playground, and I did not expect this level of accuracy!
Ilya Zakharkin retweeted
Sora is here! It's a diffusion transformer that can generate up to a minute of 1080p video with great coherence and quality. @_tim_brooks and I have been working on this at @openai for a year, and we're pumped about pursuing AGI by simulating everything! openai.com/sora
Ilya Zakharkin retweeted
Experiments with generative #AI try-on 👀
Thrilled to present what I am working on at @zero10_ar 🥼
🧵 Would you like to see how clothes look on you before buying online? Well, it may become possible very soon.
We're excited to share our latest research in #generativeAI Virtual Try-on.
Congrats to @StabilityAI and @pika_labs with recent video-model releases!
I like videos too.
Ilya Zakharkin retweeted
Check out how Cyprus-based @zero10_ar is revolutionising retail with Augmented Reality in fashion. bit.ly/3Rzq8U2
Excited to share our work on Point-Based Modeling of Human Clothing accepted to ICCV'21! 🥳
🌐 Project page: saic-violet.github.io/point-…
🎬 Video: youtube.com/watch?v=kFrAu415…
📝 Paper: bit.ly/3ax9ei6
Join us on Session 11 at iccv2021.thecvf.com on Oct 13th and 15th
Join me at 3pm (15:00 Moscow time) today at events.yandex.ru/events/chri… (in Russian)
p.s. Happy Holidays everyone 🎆
Unbelievable quality of object removal on high-resolution images. Congratulations to the colleagues! 🪄
New 🦙 LaMa inpainting is out! Trained only in 256x256 but generalizes to ~2k! The core of the system is Fourier convs in the spectral domain. It is very good at repetitive patterns!
📰 arxiv.org/abs/2109.07161
💻 github.com/saic-mdal/lama @PyTorch
🧪 bit.ly/3iPw4pv
Btw the work on point-based clothing would not exist without the Cloud Transformer architecture. @makezur you rock!
Can you get SoTA results on 4 various point cloud processing tasks with a single block? The answer is yes!
🌐Project page: saic-violet.github.io/cloud-…
📽️Video: youtu.be/lYTzLhy-ybw
🧐Paper: arxiv.org/abs/2007.11679
Join us on Session 8b at #ICCV21 on Oct 15th
Glad to share our post in Google Developers Blog 🤟
It is about a SIGGRAPH Asia 2020: Emerging Technologies project called Video-Touch arxtage.github.io/videotouch…
✋🤯 No, this isn't a sci-fi movie. You can really control a robot with your hands!
Introducing Video-Touch, the first robot-human interaction system that allows multi-user control via video calls from anywhere in the world.
See how it works → goo.gle/2UxVGOX
True AR became closer
Touch it—you know you want to. Or throw it, or dance with it, or anything else you can think of. #HoloLens2 leverages a real-time physics engine plus careful use of light and sound to replicate (and sometimes bend!) the rules of the physical world. 👽 msft.it/6013VtyWx
Check out my recent talk (RUS) on 3D clothing modeling 👘 at @phygitalism Phygital Science session ⚡️
youtu.be/ySx-iE-mb0s?t=7412
Thanks to Samsung AI Center - Moscow @SamsungRussia and @Skoltech for an opportunity to work on such a hot topic 🚀
Ilya Zakharkin retweeted
Cool neural rendering work by @ASevastopolsky from Samsung AI. Creates a 3D portrait of human head that can be relit from a video captured with smartphone. Check out the project page!
🌐 saic-violet.github.io/religh…
🎬 youtube.com/watch?v=qQnNalM9…
📝 arxiv.org/abs/2012.09963