#OldNeuroPapers #neuroscience #HistoryOfIdeas #ITcortex #ObjectRecognition

Heinrich Klüver in the 1930s wanted to study the effects of mescaline(!) in macaques after bilateral temporal lobectomy. Paul Bucy, a neurosurgeon performed the surgery and experiments. However, Bucy did not observe any of the hypothesized effects of mescaline and instead discovered that the animal had significant impairment/abnormalities. These broadly included: subdued/docile emotional expressions, adverse sexual behavior, dietary changes, utilization behaviors, increased tendency to use the mouth for exploring the world, and finally, but most importantly for our story, difficulties in visual learning and agnosia!

In the 1950s, neuroscientists and surgeons documented and confirmed similar behaviors in humans who had temporal lobectomy. Today, we know this as the famous/eponymous Klüver-Bucy syndrome due to bilateral lesions (or tumors) of the medial temporal lobe (MTL).

2/10

#OldNeuroPapers #neuroscience #HistoryOfIdeas #ITcortex #ObjectRecognition

While neurosurgeons were working with patients and characterizing the syndrome, there was a short interregnum in the world of visual neuroscience, thanks to the groundbreaking work (dare I say, paradigmatic leap) from David Hubel and Torsten Wiesel with their experiments in the striate cortex (V1, V2) and LGN.

3/10

#OldNeuroPapers #neuroscience #HistoryOfIdeas #ITcortex #ObjectRecognition

The great Karl Pribram (at Stanford) and Mortimer Mishkin (at NIH) towards the end of the 1960s performed focalized lesions in macaques in only the IT cortex (instead of the entire MTL as reported in Klüver-Bucy). In their experiments, they observed that such lesions only led to deficits in visual processing, and learning!

4/10

#OldNeuroPapers #neuroscience #HistoryOfIdeas #ITcortex #ObjectRecognition

In the middle to late 1960s, then at MIT, George Gerstein, and a dashing young Charlie Gross (my intellectual grandfather), inspired by single neuron recordings of Hubel and Wiesel, and the work of Pribram and Mishkin, stuck microelectrodes in IT cortex of awake monkeys and showed that they responded to visual stimuli. This was the first demonstration of neurons being active for visual inputs far away from the striate areas!

Later Peter Schiller joined them in the experiments. They also showed these neurons were involved in attentional mechanisms. The input stimuli, however were still rudimentary and not resembling anything “object” like: diffused light, orientation, movement etc., And then by happenstance, Charlie found "face" and "hand" cells (much like how Hubel and Wiesel found orientation cells by complete accident, slipping of the image on the projector)!

5/10

#OldNeuroPapers #neuroscience #HistoryOfIdeas #ITcortex #ObjectRecognition

Charlie Gross then moved to Princeton in the 1970s. Bob Desimone (my mentor), and Tom Albright joined as some of his first grad students. The inimitable Eric Schwartz joined them as a postdoc for a couple years. The Gross lab in the mid to late 1970s established systematically that IT neurons responded to complex visual inputs, their overall shapes, and thus objects rather than to individual features like orientations, or color, or simple curvature. This led to the funny and famous story of the "toilet-brush" neurons. The toilet-brush neurons also responded to “hand” cells that Charlie had earlier identified, so the “fingers” in the two were the commonality. This led them to come up with the idea of Fourier shape descriptors to suggest how the brain builds the "it" from the "bits" (a forerunner to all the modern linear combination of activities to give an output response, including the currently in-vogue deepnet models).

6/10