Gyro

I like capturing moments. Now I teach models to do it.
Posting every build | Growth @webuildscore

[Gyro @Gyrolens]

Testing what comes next at @webuildscore before anyone else, and this is where all of it lands.

I’m not an engineer, so you’ll see it the way most people would use it. Real runs on games, films, CCTV and sports, benchmarks against the usual way of doing things, side by side tests, and the honest time each one took.

Got something you want detected? Reply with it. The best ones get built here.

[Gyro @Gyrolens]

I’ve seen what’s behind this one. can’t wait for you to try it

[Score @webuildscore]

It learns from you now.

[Gyro @Gyrolens]

One of the things I’ve been testing at @webuildscore. Can’t say much yet, but you can see the difference.

On top, the boxes jump around and miss half of what happens between frames. Below, they just stay locked on.

It’s a small thing until you’ve watched the choppy version for a few weeks. More soon.

[Gyro @Gyrolens]

@grok you know ball

[Gyro @Gyrolens]

hey @grok remove the best vision AI platform

[Score @webuildscore]

Orange juice production line, tracked end to end 🍊

On a food production line, every unit that passes the camera matters. What goes in, what jams or falls off, and what comes out at the end usually gets checked by eye, or not at all.

Our vision models can handle it. Track every orange from the sorting line to the juicer, and any gap can be flagged before it turns into waste.

Pulp fiction, now with bounding boxes.

[Gyro @Gyrolens]

@webuildscore my goat

[Gyro @Gyrolens]

Couldn’t name myself Gyro and not do this one.

Same tool I’ve been running on CCTV and game trailers, this time on Steel Ball Run. Horses at full gallop, riders, and a stadium of people, picked up through most of the race.

[Gyro @Gyrolens]

@atifatpl Agree, and it’s only getting more true. Building the model keeps getting easier. Knowing what it should look for is the part you can’t shortcut

[Gyro @Gyrolens]

@crypt0lake brazil

[Score @webuildscore]

We got flying cars before GTA 6.

Model said "vehicle", but was wrong. Tesla Roadster it is. one sentence later it was tracking it through the whole showcase.

A first look at the new Studio. More soon.

[Gyro @Gyrolens]

been trying it here: https://t.co/4fw8YAO3os

[Gyro @Gyrolens]

Been testing the new version of Score Studio and Rockstar just handed me the perfect test footage.

Vice City has never been this tracked.

[Doodmasu @Doodmasu]

Ah, Lord Beerus, you're finally awake. Good.
Peep the Timbs.

[Gyro @Gyrolens]

@webuildscore

[Score @webuildscore]

We ran a GTA 6 "leaked" dirt race through Score Studio.

Every rider, every bike, every truck kicking up dust got picked up, the model didn’t miss a single overtake.

Rockstar, if you’re reading this, we’d like early access purely for research purposes.

[Gyro @Gyrolens]

@webuildscore casually picked up cyberleek too

[Gyro @Gyrolens]

Opened my laptop to work. Ended up running the Vehicle Detector from @webuildscore on every Rocket League replay I have. Send help, or more replays.

[Score @webuildscore]

Preparing our Security Risk Indicator model for the GTA 6 release on November 19. We're already teaching it what a robbery looks like.

Draw a box, Score Studio suggests the class, confirm, next frame. And yes, we know it's a bandana.

Try it at https://t.co/u5K9abTrCO

[Score @webuildscore]

Counting cars at the Arc de Triomphe, the roundabout with no lanes and no rules.

4 blocks in Score Studio's workflow builder: image in, vehicle detector, count, output. No code.

Swap the detector and the same 4 blocks count people, plates or anything else you point it at.

Build yours: https://t.co/lwvltsxGTm