Size and position invariance are well-established properties of object recognition. Objects presented at different retinal locations or scales elicit similar behavioural responses and neural activity patterns in object-selective cortex, and these invariances are largely explained by traditional models of the ventral visual pathway. However, much less is known about whether object recognition is also invariant to the depth cues that define three-dimensional shape. Objects can be rendered using shading, texture, binocular disparity, or structure-from-motion, yet observers readily recognize them across these diverse visual formats. Unlike size and position invariance, processing depth cues such as disparity and structure-from-motion engages not only ventral visual areas but also dorsal visual regions. Understanding whether these distinct cues converge onto a shared object representation—and how ventral and dorsal pathways contribute to this process—remains an important open question in systems neuroscience.