Build logs from active journeys, standalone tutorials, and anything else worth writing down.
60 videos. 210,000 frames. About 0.5% of them matter. This is how I got the frames out — and why I'm only saving a 400×100 pixel crop of each one.
Before you can train a model you need data. Before you have data you need footage. This is how I got it and why highlight reels beat full broadcasts for this problem.
The keychain concept started with a car and a feeling that everything in the automotive accessories market was either cheap or soulless.
The one thing you should do before every ML project. Takes 30 seconds and saves hours of dependency hell later.
I wanted my living room to react when the Leafs score. Here's the full system design and why I chose computer vision over an API.