I forced a bot to watch a stream of 1.4 million pictures from ImageNet, and then asked it to describe your selfie. The answer? "Seatbelt".
Show this thread
-
OK, I'll be the first to ask. I've only done a Kaggle "Icebergs from Space" challenge. What's the joke?... Or a reference to figure it out?... A hint please.
There is no human class in ImageNet labels, but there is a seatbelt class. Thus ImageNet-trained models tend to classify people as seatbelts (due to class co-occurrence).
-
-
Ahhh, OK. I should go explore the classic training datasets.
Thanks for your response.Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.