It could be this: http://www.labsix.org/physical-objects-that-fool-neural-nets/ http://www.labsix.org/physical-objects-that-fool-neural-nets... It is impressiv

It could be this: http://www.labsix.org/physical-objects-that-fool-neural-nets/ http://www.labsix.org/physical-objects-that-fool-neural-nets... It is impressive, but that came out a couple of days ago and is not reviewed yet, as far as I can tell. It’s not obvious from the URL but two of the authors were interns at Google, which could explain why they describe some aspects of InceptionV3 as white-box (because it's not clear if they have re-trained it themselves).