Video lecture clips: Here is a playlist of YouTube lectures that can be useful (lecture clips from short to long, so you can choose which ones suit you) ,

The title and abstract of the lectures is:

Title: From V1SH to CPD in the Encoding-Selection-Decoding (ESD) framework for understanding vision.

Abstract: V1SH is the V1 Saliency Hypothesis that the primary visual cortex, V1, creates a bottom-up saliency map to guide gaze to salient locations so as to devote the limited brain resources to scrutinizing the visual input within the attntional spotlight. CPD is the central-peripheral dichotomy that central and peripheral vision differ such that top-down feedback from higher to lower visual cortical areas to aid visual recognition is mainly directed to the central rather than the peripheral visual field (and this central-peripheral difference is in addition to their difference in visual resolution). In the ESD framework that vision is composed of encoding, selection, and decoding stages, encoding means to represent visual inputs optimally, e.g., for better information representation efficiency; selection means to select only a tiny fraction of visual input into the attentional or information bottleneck for deeper processing, and decoding means to infer visual scene properties (e.g., a face identity) from the selected visual inputs. In this framework, V1SH pertains to the selection stage, and implies that the attentional bottleneck starts at the output of V1 to higher visual areas. CDP pertains to the decoding stage, is in light of the selection behavior, and is also motivated by experimental evidence for V1SH. I will explain in detail the computational formulations, theoretical predictions, and experimental evidences for V1SH and CDP, and relate them to visual behavior and phenomena such as eye movements and visual illusions. Recommended preparation materials for this lecture can be obtained from https://lizhaoping.org/zhaoping/CNeuro2020_teaching.html