Videos footage from RuCCS Colloquium Talks can be found on the RuCCS YouTube Channel. For all other events, please check the sponsor's website for more detail.
To filter by event category, click on the event category link in the table below or use the menu on the right.
List of Past Events
What Neurons in Monkey Inferotemporal Cortex Tell Us about Human Perception
Dr. Carl Olson
Monday, March 08, 2010, 12:00pm - 07:00pm
Carnegie Mellon University, Center for the Neural Basis of Cognition
Our ability to recognize visual images depends critically on inferotemporal cortex (IT). Neurons in IT, when tested with any arbitrary set of images, are like snowflakes: all alike at one level (in that they exhibit pattern selectivity) each unique at another level (in that each responds to a different subset of images). While the fact of image selectivity is well established, the underlying principles are still poorly understood. I will describe experiments that cast light on the underlying principles by showing that IT neurons are very sensitive to some attributes of images and are quite insensitive to others. I will further show that these results have direct parallels in human visual performance. An example is given below.
To human viewers, the asterisk stands out unmistakably from the surrounding rectangles (A). The same is true for neurons in IT. A representative neuron responds strongly to the asterisk and not the rectangle (B). In a population of around a hundred neurons, the two forms elicit markedly different firing rates (C). However, when the two forms are combined into a compound image, the arrangement with asterisk above does not stand out perceptually from the arrangement with asterisk below (D). The same is true for neurons in IT. A representative neuron responds at similar rates to the two images (E). The population as a whole barely discriminates between them (F). This indicates a severe limit on the ability of observers and neurons alike to detect differences between images based solely on the global arrangement of local elements.