iiinova — where your digital self keeps evolving

Vahe Taamazyan @vaheta.bsky.social · 06/01/2026

We’re looking for someone with experience in: - Model Distillation - Quantization - Deep Learning Optimization 2/3

Vahe Taamazyan @vaheta.bsky.social · 06/01/2026

🚨I’m looking for a talented intern to join me at Intrinsic this summer! Our team is focused on a hard problem: ensuring high-performance 3D perception models can run efficiently on constrained hardware to enable real-time robotic decision-making. 1/3

Vahe Taamazyan @vaheta.bsky.social · 02/07/2025

There’s definitely a connection

Vahe Taamazyan @vaheta.bsky.social · 13/06/2025

Pretty sure my first words were cv2.imread. or maybe cv::imread, now that I think about it. @opencv.bsky.social @philnelson.bsky.social

242

5/ Check out the video in the first post of the thread to see one of the submissions in action. The winning teams will be announced at the CVPR 2025 Perception for Industrial Robotics Automation Workshop. Stay tuned!

000

4/ Some stats: - 450+ teams registered - Dozens of teams invested hundreds of hours - Top 5 submissions were deployed on a real robot using Intrinsic Flowstate + ROS We tried to pick parts based on participants' predicted poses - live!

100

3/ This is one of the toughest problems in computer vision, and it’s still far from “solved” - especially at the levels of accuracy and robustness required in real industrial settings.

2/ Participants were given multi-view, multi-modal images and tasked with training models that not only detect objects, but also predict their full 3D pose. That means 3D position + rotation - exactly what a robot needs to grasp and manipulate objects accurately.

100

This has been one of the most exciting and demanding projects I’ve worked on. Over the past 6 months, we teamed up with OpenCV, the BOP benchmark, and the CIRP Lab @uhmanoa.bsky.social to run a global challenge on 6DoF pose estimation for complex industrial parts. 🧵

Vahe Taamazyan @vaheta.bsky.social · 01/04/2025

Want to know if humanoid robots have finally become practical? See if more of them roll smoothly on wheels instead of stumbling around on legs. Until then, they’re stuck in demo mode - not deployment and scaling mode.

Vahe Taamazyan @vaheta.bsky.social · 19/03/2025

Gotta admit – I enjoy deleting code way more than writing it.

Vahe Taamazyan @vaheta.bsky.social · 14/03/2025

Ok, someone literally called their paper Foundation X: arxiv.org/pdf/2503.09860

arxiv.org

Vahe Taamazyan @vaheta.bsky.social · 12/03/2025

Having your own pattern is a nice flex, so I’m happy for Bowen actually! :D

000

Vahe Taamazyan @vaheta.bsky.social · 12/03/2025

I only know Bowen Wen’s Foundation Pose and Stereo papers that follow that pattern. What else is there?

100

Vahe Taamazyan @vaheta.bsky.social · 12/03/2025

CV/ML paper naming strategies: - Answering yes to an obvious question? → “Do You Need X for Y?” - New one-shot model? → “X Anything” - Overhyping something trivial? → “All You Need Is X”

Vahe Taamazyan @vaheta.bsky.social · 14/02/2025

Find more details and apply here: 🔗 job-boards.greenhouse.io/intrinsicrob... 🔗 job-boards.greenhouse.io/intrinsicrob...

011

Vahe Taamazyan @vaheta.bsky.social · 14/02/2025

🚀 Internship Opportunity! 🚀 Our small research team is looking for two talented interns to join us! If you're passionate about cutting-edge research at the intersection of computer vision and robotics, this is your chance to contribute to an exciting, high-impact project.

141

Vahe Taamazyan @vaheta.bsky.social · 06/02/2025

Prediction: There’ll be more people calling themselves software developers in 2–3 years than there are now.

000

I’m wondering how long it’ll take until people start memorizing openings for all 960 starting options

I made a bet with a friend on 1 Nvidia stock that in 5 years, I’ll be able to buy a robot that can cook pasta in my kitchen, straight from raw ingredients, with no special setup. Just in case, I’ve already bought the stock😆. What are my chances?

000

And in million years you’ll get the answer 42

Here is the challenge website: bpc.opencv.org

000

The Bin Picking Challenge is live! Detect the 6DoF poses of 10 challenging parts and compete for $60K in prizes. Some of them are quite challenging to detect accurately. The video shows a few of the parts in rotation - are you up for the challenge?

Vahe Taamazyan @vaheta.bsky.social · 01/02/2025

The Bin Picking Challenge is now open!

040

Vahe Taamazyan @vaheta.bsky.social · 29/01/2025

The only reasonable explanation I can think of is that Nvidia’s moat is larger in training than in inference. If training becomes much more efficient, demand saturation becomes a risk. But even this explanation feels somewhat weak.

000

Vahe Taamazyan @vaheta.bsky.social · 26/01/2025

I see people debating the Humanity’s Last Exam dataset. I actually contributed a problem to it. For the past 2 years, I’ve used a much simpler version of that problem to evaluate new LLMs—and they’ve all failed miserably. I can’t speak for other problems, but once an LLM solves mine, I’m retiring 🫡

160

Stay tuned for updates! Follow OpenCV’s official channels and my account for announcements, resources, and tips. Looking forward to seeing your innovative solutions! 5/5

000

Intrinsic is a software and AI robotics company at Alphabet. They’re building a platform for the next generation of intelligent automation.

Organized by OpenCV, with sponsorship from intrinsic.ai and Orbbec, plus support from BOP (bop.felk.cvut.cz) and the Computational Imaging and Robotic Perception Lab at University of Hawaii at Manoa. 4/5

intrinsic.ai

Intrinsic

100

The challenge focuses on 6DoF pose estimation for challenging parts. It's a real-world problem with real impact. The final stage even features a real robot! (See the attached GIF above for some detected object poses in action!) 2/5

100

Vahe Taamazyan @vaheta.bsky.social · 02/12/2024

Big news for computer vision enthusiasts! Intrinsic.ai is sponsoring the Bin-Picking Challenge for Perception, with $60,000 in prizes and a chance to present your work at the CVPR Perception for Industrial Robotics Automation Workshop! 🧵 1/5

130

Mitsuba is amazing, thank you for your work!

Well, my understanding is that Bluesky is an open platform that enables this kind of data scraping by design, so I disagree with you and with all the people attacking Daniel. For those not liking it there is X/twitter. But I don’t really feel inclined to argue further about it.

Hey Daniel, I don't know you personally, but I'm so sorry to see you going through this. Sending you strength and support.

2140

Thank you, we'll check the problem

Huh, could you please tell me which version of Chromium you're using? Thanks for catching that issue with Safari – we'll fix it! One of the downsides of working at a Google-related company is that it's easy to forget about other browsers, which we shouldn't.

Is this the reason why people do LayerNorm before the MHA, nowadays? 😄

000

A generalizable, Al-enabled 3D vision system for industrial robotics

4/ My colleagues will present this work next week in Tokyo. 🇯🇵 If you’re attending SIGGRAPH Asia, don’t miss it! 🔗 Intrinsic Blog Post: intrinsic.ai/blog/posts/a... 🔗 Paper page: akasha-imaging.github.io/plenoptic-vi... 🔗 Paper pdf: akasha-imaging.github.io/plenoptic-vi...

intrinsic.ai

At SIGGRAPH Asia we are presenting a plenoptic 3D vision system that enables intelligently adaptive robotic solutions that also generalize across applications

251

3/ Our plenoptic 3D vision system solves this by significantly improving both reliability and accuracy in these challenging environments. This approach opens new possibilities for robots to function more effectively across a variety of scenarios.

2/ Robots rely on precise 3D scene data to operate effectively, but traditional 3D cameras often fail with transparent or reflective surfaces - a big problem in industrial settings.

Vahe Taamazyan @vaheta.bsky.social · 25/11/2024

1/ Excited to share that our latest work from Intrinsic will be presented as a paper at SIGGRAPH Asia 2024! 🎉 We’ve developed a plenoptic 3D vision system that addresses a key challenge in industrial robotics: providing robots reliable 3D input data. 🧵⬇️

3212

"Surely You're Joking, Mr. Feynman!" and ""What Do You Care What Other People Think?: Further Adventures of a Curious Character" are highly recommended for anyone planning to do science.

000

Vahe Taamazyan @vaheta.bsky.social · 25/11/2024

My feelings about humanoid hype