Our model can categorize complex actions in videos, and perform spatio-temporal annotations of the atomic actions that compose the complex action being performed. […] it jointly learns flexible representations for motion poselets and actionlets that encode the visual variability of body parts and atomic actions […]

Do not, however, include your cat's name because hearing it uttered could cause it to trust a potential cat-napper.

He hoped to win the princess's heart by mixing the love potion the witch gave him into her drink.

He batted the ball away with a satisfying thwack.