Can a deep net see a cat? – Piekniewski’s blog

Can #deeplearning recognise a cat?  The answer is not obvious. #AI #vision

  • In summary, the thesis of this post is that visual perception is NOT a solved problem and requires more work, likely a fundamental shift.
  • An adversarial example constructed from the category “Egyptian cat” classified with 98% confidence.
  • BTW: since I uploaded these cat images to ClarifAI, they may become part of their training set future results from them might be different than those presented .
  • The “best cat” is quite confidently classified as a dalmatian (0.74) and only with 0.04 confidence as the “Egyptian cat”.
  • It does classify my best cat as a cat (success!

In this post I will explore the capabilities of contemporary deep learning models on the vitally important task of detecting a cat. Not an ordinary cat though, but a sketch of an abstract cat. This task matters because success tells us something about whether a visual system has learned generalization and abstraction  — at least on par with a 2-year old. This post is inspired by my ex co-worker Peter O’Connor who tried similar experiments on LeNet several years ago. In addition, this post is a continuation of this blog’s highly popular “Just how close are we to solving vision?” which to-date has amassed nearly 15,000 hits. Let’s begin by introducing my menagerie:
Continue reading “Can a deep net see a cat? – Piekniewski’s blog”