In 1953, Barlow discovered that the frog brain has neurons which fire in response to specific visual stimuli. This was a precursor to the work of Hubel and Wiesel on visual receptive fields in the visual cortex. He has made a long study of visual inhibition, the process whereby a neuron firing in response to one group of retinal cells can inhibit the firing of another neuron; this allows perception of relative contrast.
In 1961, Barlow wrote a seminal article[4] where he asked what the computational aims of the visual system are. He concluded that one of the main aims of visual processing is the reduction of redundancy, which has been extended to the efficient coding hypothesis.[5] While the brightnesses of neighbouring points in images are usually very similar, the retina reduces this redundancy. His work thus was central to the field of statistics of natural scenes that relates the statistics of images of real world scenes to the properties of the nervous system.[citation needed]
Barlow also worked in the field of factorial codes. The goal was to encode images with statistically redundant components or pixels such that the code components are statistically independent. Such codes are hard to find but highly useful for purposes such as image classification.[citation needed]
Barlow was married twice and fathered seven children.[1] In 1954, he married Ruthala Salaman, daughter of M.H. Salaman. They had four daughters: Rebecca, Natasha, Naomi and Emily. They were divorced in 1970. In 1980, he married Miranda, daughter of John Weston-Smith. They had one son, Oscar, and two daughters, Ida and Pepita.
