📯CALL FOR PAPERS: Submit work to the Workshop on Computer Vision for Mixed Reality @ CVPR 2025!
📆 Deadline: March 28 → Papers appear in the CVPR Workshop Proceedings
✨ Topics: 3D/Image/Audio Capture/Understanding/Reconstruction, Spatial AI Agents, Human-AI Interaction.
cv4mr.github.io
Key roles are around mobile and web development, UX design, 3D computer vision, and image-based machine learning. If you’re looking for a change, reach out and let me know how you see yourself contributing!
📈 For the upcoming year, my allegiance will now be with the Steelers, as here in 2025 I'm building a team of smart creatives in Pittsburgh to lead innovation for SR Measure and Stockpile Reports.
Eagles win! 🦅
Of course I had to honor them with a stockpile, given that I spend most of my time thinking about ways to improve or invent stockpile measurement techniques via 3D computer vision and machine learning. 😄
Getting ready for something exciting...
📈 Here in 2025 I’m building a team of smart creatives in Pittsburgh to lead innovation for SR Measure and Stockpile Reports, with key roles around mobile and web, UX design, machine learning, and 3D computer vision. If you’re interested, reach out and let me know how you see yourself contributing!
🎨 Generative AI was used as effective entertainment. During one of the intermissions, random fans in the crowd were turned into bobblehead characters. It was fun and well done!
Let’s Go Pens! 🐧
🚀 That puck moves fast! With shots traveling 25-45 m/s, I wonder what an augmented reality UI could look like that leverages event cameras (cameras that offer high-speed, asynchronous, per-pixel updates for intensity changes). Though, the camera might track the puck faster than our eyes can keep up!
🏒 Hockey + AI ?
I attended my first @penguinsnhl.bsky.social hockey game this weekend - it was a lot of fun!
Two things that surprised me:
📈 In 2025 I’m building a team of smart creatives in Pittsburgh to lead innovation toward the goal of digitizing every pile on the planet every second of every day – reach out if you see yourself contributing!
🔬 I’ve used hillshade before to emphasize elevation changes in terrain, and so I could mentally visualize what it would look like to turn it into an animation. However, nothing beats actually doing the experiment and seeing the results.
💡 Sometimes I get a random idea that I just have to try, either because I think it sounds fun, because I think it could help our business, or because I think I might learn something new (or more typically it’s all three!).
🤔 What does it look like when you continuously vary the sun azimuth in a hillshade visualization of a stockpile site?
That's an awesome setup!
🚀 If you're looking to make a change and want to join a team of smart creatives who aren't afraid to take on challenging problems, reach out and let me know how you see yourself contributing!
🏢 Here in 2025 I'm opening an office in Pittsburgh and building a team to tackle and lead new innovation for SR Measure and Stockpile Reports, with key roles around mobile development, UX design, 3D computer vision, and image-based machine learning.
📏 Can you spot the tape measure that was being used as a test to manually verify the scale?
❄️ This past week I got out of my office and measured some snow-covered piles (one of the environmental challenges of being in Pittsburgh over the winter).
📱 The animation above shows a sparse point cloud that was computed on an iPhone 16 within seconds for one of those piles.
❓ The answers aren't in the office...
🥾 One of the core principles that we believe in at EveryPoint.io is that we need to get out and get our boots muddy in order to truly understand our customers and build a better product.
Hope you feel better soon!
Slightly related: During my holiday break I really enjoyed reading (or re-reading) How Google Works. Lots of inspiring thoughts for how to build a team, product, and company.
🏝️ Benefit of vacation: It’s important to get away and do things other than work.
Downside: When writing my first line of C++ for 2025 I tried to end it with a period (as if it were a sentence) instead of using a semicolon. 😬
New to Bsky and interested in #computervision? Check out my 2 Starter Packs of #computervision researchers.
go.bsky.app/M7HGC3Y
go.bsky.app/Gjcs339
I love it!
Beautiful results! Nice work! 🎉
#3DV2025AMA Fourth guest on the Ask Me Anything series:
Jia-Bin Huang @jbhuang0604.bsky.social from University of Maryland & Meta! Jia-Bin was the engine behind all the fun tweets of last year 3DV 🌟
🕒 You have now 24 HOURS to ask him anything — drop your questions in the comments below!
I know I'm past the 24 hour limit, but @jbhuang0604.bsky.social, what's your favorite part of the work that you're doing now and/or which of your past papers was the most fun to work on and why?
RANSAC is a powerful tool. There may be more up-to-date resources out there, but for me, the USAC paper by Rahul Raguram et al. is a great overview of the many kinds of modifications or improvements that can be made:
rahulraguram.com/assets/pdf/R...
Welcome! 🎉
Because of this release I just learned about PoseLib. It looks like a nice C++ library for pose estimation!
github.com/PoseLib/Pose...
