Attention Conservation Notice : Slide decks from my two talks; if you didn’t see my talks these will likely be of limited interest.

Slides from my BSides talk – “Security Data Science: getting the fundamentals right” – are here. It’s worth noting that this talk is to some degree aspirational; I don’t think I’ve ever managed to run a project where we did all of them at once, but we’ve done all of them individually and the projects tend to run better in direct proportion to how many of them we actually do implement.

Slides from my AI Village @ DEF CON talk – “Hacking Facial Recognition Systems” – are here. Note that the scope I looked at was limited, the talk focuses on the “adversarial ML” side of things; you should also check out @adversariel’s talk on attacking entire systems “Don’t Red Team AI like a Chump”. I am, arguably, red-teaming like a chump :)

Please note that the Colab notebook linked at the end of the facial recognition slide deck is not guaranteed to stay up indefinitely; I’ll probably move it to GitHub at some point and will post here when I do. Also apologies for the really janky optimization loop and loss function for the white box attack.