Part 2: https://r2d3.us/visual-intro-to-machine-learning-part-2/

Thanks for sharing. I went through a similar process recently, manually stitching together fire insurance maps from my hometown. A large part of the business district burned down in 1916, but maps from a couple of years earlier survived. It worked, but it was tedious. This would have saved a lot of time. Here's the link if anyone wants to check it out: https://fire.gorch.com/

Thanks for sharing, this is really nice. I've added it to the list of other projects I have on the site [0]. Feel free to get in touch if you want to use OIM for other years of Paris, TX.

[0] https://oldinsurancemaps.net/more-about-sanborn-maps/

Oh, cool, Thanks so much!

Timing of your post is spot on. I just emptied a drive to prep for a Linux switch this morning ( for the same reasons ) :D

Love it

Listened to Flood continuously in high school. In many ways, I felt it empowered me to embrace my own weirdness.

"Was it love or was it the idea of being in love?" - "One Slip", Pink Floyd

This great, Thanks for sharing!

I wonder if Regenerative AI would be a more suitable name.

Creator here. I add the static to mask the video buffering ( since each channel change triggers a video load ) I'm flattered though that you think of me as young. I very much was a child of the '80's :D

You could add a "high bandwidth & low latency" mode, where when active, you load the current video + the next one. So when the user goes to the next channel, it's already playing but muted and not visible, and you start playing the next-next channel hidden again :)