Big news! At www.beisbol.ai we proudly present our Framing App! State-of-the-art object tracking for quantifying and analyzing catchers' framing skills! Measure distance, displacement, velo and multiple other metrics to gauge the framing process and extract the needed data for further eval.

I'm thrilled to announce that we're officially launching beisbol.ai!

For years, we have been passionate about the intersection of baseball and artificial intelligence. Today, we're taking a huge step forward in our mission to revolutionize the game.
#beisbolAI

And finally trims the frames that are not part of the detected motion.

In this case, the clip was trimmed to a little over one second using the tool. This can be done in batches to process multiple videos and then concatenate them together for other processing purposes (8/n)

We obtain the full segmented video (6/n):

There are multiple paths to working this out but at the core, you want to: find the pitcher, track its motion from start to end (ball release), trim clip accordingly.

Let's use this pitch from BBB, as an example. It's a typical 6-seconds short clip from Savant but there (3/n)

Let's discuss setting up Computer Vision pipelines for your baseball needs, and let's use a practical example: an AI Video Editor, which at BaseballCV, is one of the projects we are currently working on.

If you used our app, baseballcv-savant-data-and-video.streamlit.app, for creating this (1/n)

Spencer Strider is looking more and more like Quadzilla again!

Watch him making the Mets swing and miss in this highlights reel!

You can create yours easily using BaseballCV’s Savant Data and Video app:

baseballcv-savant-data-and-video.streamlit.app

It’s easy and requires no login!

And you get:

Do you wanna see Mike Soroka mowing down hitters left and right? Try Highlights but for pitchers:

Which results in:

Let's look at an example:

Searching for highlights from Ronald Acuña Jr. in the last 10 days.

that it can be done, needing only a reasonable quality video.

That's what our CommandAnalyzer class, part of the Baseball Tools in BaseballCV aims to achieve, after these proof of concepts.

It'll be available soon from github.com/dylandru/Bas...

but the timing varies greatly; that's an issue. Also, some barely show intent location and then "hide" the glove, so these two things are opposing in how to handle them and mess up with each other, for example. There are a lot of things to take care of but the point is (4/)

Trying to catch, pun intended, the Target position of the glove, just by using the full sequence's glove position tracking, is not trivial as there are many implicit variables working at the same time.

A lot of catchers put the glove down after showing the intended location (3/)

After Dylan Drummey and I got a fully working catcher's glove tracking tool, we wanted to find useful applications for it and this is one that popped up immediately: Can "intend/actual" tracking be done by means of using only that glove data?

Well, it looks like it does. (2/)

How complicated is it to calculate any sort of deviation between where a pitcher "Targeted" to throw a pitch and where it "Actually" landed?

It can be very complicated, especially if you don't use sophisticated systems involving advanced tracking devices.

Or maybe not? (1/)