Satya Mallick
3.4K posts
Satya Mallick
@LearnOpenCV
CEO, https://t.co/CzUdJlxzJM. Course Director, https://t.co/O2Tz9vUOQ8 Entrepreneur. Ph.D. ( Computer Vision & Machine Learning ). Author: https://t.co/olraDEG5Ue
San Diego, CA Katılım Haziran 2008
903 Takip Edilen14.1K Takipçiler
No, they don't.
Mistakes like this do not happen by chance.
They took a calculated risk that people will be okay with such a violation of privacy.
Show them it is not acceptable.
Stop using xai
SpaceXAI@SpaceXAI
We care deeply about your privacy and respect customer choice. For teams using zero data retention, no trace and code data is ever retained. All API key use of Grok Build also respects ZDR. If ZDR is disabled, the /privacy command is available in the CLI to disable data retention, which also deletes previously synced data. Run the /privacy command to view or change your settings at any time.
English
(OpenCV 5: From Contours to Vision Language Models)
OpenCV used to mean read images, find contours, draw boxes. OpenCV 5 runs vision language model inference end to end, with a sample running PaliGemma 2 on the DNN engine.
Full breakdown ⬇️
opencv.org/opencv-5/
English
(OpenCV 5: A Cleanup and a Serious Move Toward Modern AI)
OpenCV 5 is not a flashy release. It is a cleanup and a serious move toward modern AI: new DNN engine, ONNX coverage from under 23 percent to over 80 percent, and 3D vision split into its own modules.
Full breakdown ⬇️
opencv.org/opencv-5/
English
(How Computer Vision Solved the Greatest Soccer Mystery of All Time)
1996, Wembley. The most argued-about goal in football history. Zisserman and Reid rebuilt the ball's trajectory from unsynced broadcast footage and found it fell short by at least 6 centimeters.
Full deep-dive ⬇️
learnopencv.com/how-computer-v…
English
(Semi-Automated Offside Technology, Part 2: Cameras, Digital Twins, and the Offside Line)
The ball answers when. The cameras answer where. The digital twin answers which body part. The officials still answer the football question.
Full deep-dive ⬇️
learnopencv.com/world-cup-2026…
English
(Semi-Automated Offside Technology: How the Ball Times the Touch)
The World Cup ball doesn't decide offside. It timestamps it: ~500 readings a second from an IMU inside, pinpointing the exact touch.
Full deep-dive ⬇️
learnopencv.com/world-cup-2026…
English
The whole AI field in one sentence: machine learning is how it learns, deep learning is the brain, computer vision the eyes, NLP the ears and mouth, robotics the hands and feet, recommendation systems the crystal ball, and fuzzy logic the conscience.
Learn the vision side with code: learnopencv.com
English
Happy 250th Birthday, America!
Before I came to America, I knew it as a land of opportunity - a place where hard work is celebrated, failure is tolerated, and enterprise is rewarded.
But after I arrived, I discovered something even more remarkable, something you can only truly experience up close: the warmth and generosity of its people.
America is not just a country of opportunity. It is a country that welcomes.
Thank you, America.
English
2/2 Link to the article I wrote about Semi-Automated Offside Technology (SAOT).
learnopencv.com/world-cup-2026…
English
1/2 New article about Semi-Automated Offside Technology (SAOT).
Learn about
1. Multi-camera player tracking 3D pose estimation and digital twins
2. Connected-ball sensing with IMU-style accelerometer and gyroscope data
3. Sensor fusion between ball timestamps and player geometry
4. Why VAR and referees are still part of the decision loop
English
The Portugal versus Croatia game was decided by this ball technology today. Brutal.
Satya Mallick@LearnOpenCV
(Inside the World Cup 2026 Ball, Part 3) The Trionda is not a soccer ball with a chip inside it. It is a real-time sensor platform reading acceleration and rotation 500 times a second, feeding touch, offside, and handball decisions. Full deep-dive 👇 learnopencv.com
English
(Inside the World Cup 2026 Ball, Part 3)
The Trionda is not a soccer ball with a chip inside it. It is a real-time sensor platform reading acceleration and rotation 500 times a second, feeding touch, offside, and handball decisions.
Full deep-dive 👇
learnopencv.com
English
Finally, I have the updated Siri and it is pretty functional.
Nothing magical but not having to click an extra button to get to a voice assistant is pretty nice.
English
Y Combinator makes something people who make something people want want.
English
"Am I too late for an AI career?" No.
We get ~1,000 applications per AI role and reject ~975. That is not brilliant competition, it is shallow resumes. And entry-level AI still pays ~$120K, which means qualified people are still scarce.
Build a strong, wide foundation 👇
learnopencv.com
English
(Inside the World Cup 2026 Ball -- Part 2)
The hard part of the World Cup 2026 ball was not the sensor. It was hiding a full computer inside it without changing how it flies.
Processor, battery, transmitter, all balanced with counterweights so the ball does not wobble.
Embedded IMU on a real vision device 👇
learnopencv.com/depthai-pipeli…
English
(Inside the World Cup 2026 Ball -- Part 1)
The most important motion sensor at the World Cup 2026 is not a camera. It is the ball.
The Adidas Trionda runs a 500 Hz IMU: one motion reading every 2 milliseconds, measured from the ball's own movement.
How we track a ball from the outside, with cameras and Kalman filters 👇
learnopencv.com/the-complete-g…
English