11. Adaptive retina pixel circuit by Carver Mead (Adapted from [4]) Zaghloul and Boahen improved the silicon retina by Mahowald and Mead, by adding both sustained (parvo) and transient (magno) cellular pathway functionalities [33, 34]. This simplified model of all five layers of retina comprised of both outer and inner retina functionalities. The inner retina model added the functionality of contrast gain control and the outer retina model could effectively carry out spatio-temporal bandpass filtering and local gain control. This model exhibited biological retina-like ability to adapt to varying light and contrast conditions and dynamically realign spatial and temporal filtering. The 3.5 × 3.3 mm 2 chip they designed could accommodate 5760 phototransistors at a density of 722 per mm 2 and 3600 ganglion cells at a density of 461 per mm 2 . This implementation provides an output of spike trains which is analogous to ON and OFF centre wide-field transient and narrow-field sustained ganglion cells. This approach emphasised extensively on modelling biological retina closely, but the large mismatch, and 1-2 decades of Fixed Pattern Noise (FPN) restricted its application to laboratory experiments.