OpenCV - Open Source Computer Vision Library

The world's largest computer vision library, used in an estimated 80% of embedded vision projects and actively developed since June 2000. Non-profit.

[Paschaline Agughala | K–12 Coding & Robotics Tutor @Pct_unplugged]

I created a Gesture Visual project using @python @opencvlive , Mediapipe, Numpy. had so much fun doing this #python #ComputerVision #Learning #EdTech #AI #GestureControl

[Jordi G. Castillón @gcjordi]

My project for the OpenCV AI Competition 2026: CiberIA VisionGuard — cognitive security for Agentic Vision, powered by OpenCV 5 and AWS. https://t.co/d0OBbslCKf @opencvofficial @awscloud @devpost

[OpenCV - Open Source Computer Vision Library @opencvofficial]

Live at 9am Pacific, we'll sit under the learning tree with a CVPR highlighted paper, direct with one of the primary authors. Tune in!

[OpenCV - Open Source Computer Vision Library @opencvofficial]

Thursday: Agastya Kalra from Intrinsic (Google) presents 3PT.

Detection, segmentation, and 6DoF pose in two RGB-only multi-view transformers. First place in both the Industrial Robotics and AR/VR tracks of BOP 2025.

Watch: https://t.co/QZ4y6ZducU

[OpenCV - Open Source Computer Vision Library @opencvofficial]

Thursday: Agastya Kalra from Intrinsic (Google) presents 3PT.

Detection, segmentation, and 6DoF pose in two RGB-only multi-view transformers. First place in both the Industrial Robotics and AR/VR tracks of BOP 2025.

Watch: https://t.co/QZ4y6ZducU

[Open Robotics @OpenRoboticsOrg]

We've extended regular price ROSCon Global registration until this Thursday, August 27th at 11:59pm PST thanks to the support of our colleagues at @SeeTorontoNow!

💰Secure your tickets before the deadline to save $300 US / $416 CAD!

🇨🇦Looking forward to seeing you in Toronto!

[Kat Scott 🐀 @kscottz]

Looking for American made robotics hardware for Physical AI?

I'm helping my colleagues from the Midwest throw a party in the Mission.

We'll also be talking about open source, ROS, and the open source Physical AI ecosystem.

Should be fun.

[OpenCV - Open Source Computer Vision Library @opencvofficial]

An OpenCV AI Competition win is a resume line that means something — projects judged by an independent panel from industry and academia. Past winners built assistive wheelchairs, STEM tools, and ag-tech. Your turn: https://t.co/trYUOksTHc

[Satya Mallick @LearnOpenCV]

I have certainly experienced this, but I cannot confirm whether my usage has just shot up or they have throttled limits.

I have a feeling that these companies used to nerf the quality of the models before.

Now they are keeping the quality the same but throttling the opaque limits.

[Kat Scott 🐀 @kscottz]

I watched a great talk by @zipline at @foxglove Actuate yesterday about the amount of work and testing they've done to mitigate noise from delivery drones and also prevent screw ups like this one.

Big companies can't compete with a srartup determined to succeed.

[OpenCV - Open Source Computer Vision Library @opencvofficial]

We're live, pal!

[OpenCV - Open Source Computer Vision Library @opencvofficial]

Where does it go next?

Jason Ren of @allen_ai previews MolmoMotion: forecasting point trajectories in 3D from a plain-language instruction — built on Molmo's pointing + tracking.

Live demos, open weights, open data, open Q&A.

🗓 08/20 @ 9am PT
Register: https://t.co/kgPvCFhKFb

[Luxonis | Robotic Vision @luxonis]

Is your line moving at high speeds? What about varying code types? Different package surfaces or textures? Different working distances?

The Luxonis OAK 4 line, integrated with the Luxonis Barcode Reading Solution, takes all this guess work out of your operation - ensuring that these variables are not the reason you have critical missed reads in your environment.

Book time with our sales team to discuss piloting our solution today! https://t.co/UZxCAEI7MU
Learn more: https://t.co/JN76oX7LVs

[Simone Andreotti @Simone_dev_com]

Stepping inside a retro pixel portal in the real world. Built with Three.js, 8th Wall, and custom GLSL dithering shaders running on mobile Safari. #creativecoding

[Satya Mallick @LearnOpenCV]

How the Swarm Checks Its Own Work

The secret to my content pipeline is not one smart agent. It is a swarm where one agent builds the asset and another agent verifies it, then sends it back to be recreated if it fails. No human in the loop.

If you have built with agents, tell me your story below.

[Andrea @acolombiadev]

I work at GitHub. yesterday was rough and i'm not pretending otherwise. full root cause report is up if you want the timeline and numbers, and what we are doing to prevent this from happening again.
https://t.co/05do2WFoMa

[Roboflow @roboflow]

We are hosting the Visual Intelligence Summit on October 22 in SF.

The physical world is AI's biggest opportunity and the best way to advance humanity in our lifetimes. Few people are building for AI beyond the screen.

The Summit is for the people who see it differently and want to be at the frontier of bringing AI into the physical world.

If you’re building for that future, join us.

[OpenCV - Open Source Computer Vision Library @opencvofficial]

@RenZhongzheng See you on Thursday!

[Jason Ren @RenZhongzheng]

I'll be giving a talk about our rencent work MolmoMotion at OpenCVLive!
Join me on Thursday, August 20 at 12pm ET / 9am PT. Super excited!!

[OpenCV - Open Source Computer Vision Library @opencvofficial]

Where does it go next?

Jason Ren of @allen_ai previews MolmoMotion: forecasting point trajectories in 3D from a plain-language instruction — built on Molmo's pointing + tracking.

Live demos, open weights, open data, open Q&A.

🗓 08/20 @ 9am PT
Register: https://t.co/kgPvCFhKFb

[Satya Mallick @LearnOpenCV]

I have been working on a video generation project that converts my blog posts into a video walkthrough.

My current setup usually solves it in 2-3 shots for medium sized posts with 5-10 hours of thinking time on Codex 5.6 Sol + Ultra.

I tried it on a more complex post on YOLO26 segmentation fine-tuning and it produced mediocre results.

After a few trials it was giving okish results.

Finally, I gave it detailed feedback on what needed to be done to make it publishable.

It worked on the project for 33 hours.

Result?

Utter garbage!

Taming the AI beast on subjective tasks is quite a mystery.

[OpenCV - Open Source Computer Vision Library @opencvofficial]

Build computer vision that solves an actual problem — healthcare diagnostics, accessibility, industrial inspection, autonomous systems, or otherwise, it’s your call. OpenCV AI Competition 2026, powered by AWS. $12K in prizes. Register now: https://t.co/trYUOksTHc