Skip to content

#computer-vision

2 posts

Tags
  1. 005

    One body, two heads: training a HydraNet that draws boxes and masks, in the browser

    Emoji fruit as training data, and a small network with one shared trunk and two output heads, trained from scratch in your browser. It finds the fruit within ten seconds. Then we measure honestly what multi-task learning bought us and what it cost.

    Interactive#computer-vision#multi-task#from-scratch

    10 min
  2. 001

    A CNN from scratch: watching a convolutional network see, in your browser

    A handwritten-digit classifier written in plain TypeScript with no machine-learning library, then opened up so you can look at the output of every layer.

    Interactive#computer-vision#cnn#from-scratch

    7 min