Abstract
We take for granted that when we open our eyes we experience a rich and complex scene, extending over a wide field of view, an experience described by William James as a stream of visual consciousness. But what we see is the result of integration of incoming information over time and space. Recently scientists have discovered interactions also between foveal and peripheral visual information, using different methodologies. Here we bring together the study of some phenomena originally described as illusions (Uniformity illusion; Honeycomb illusion) with accurate psychophysical measurements of sensitivity to visual form in the periphery, and we study clinical cases of patients with central vision deficits. Central vision is critical for shape perception, and contributes also to the analysis of shape information in the periphery. One situation is when extrapolation and filling-in processes lead to properties from the centre to be perceived as extending in the periphery. We study systematically the conditions of continuity between regions that allow such extrapolation. We predict that information about discontinuity between regions will modulate their interaction. This is tested in a series of experiments (PD1) for 2D shapes and in a second series (BO2) for solid shapes and illumination. Besides having an immediate effect on peripheral appearance, foveal stimulation has also longer-lasting effects on visual processing, e.g., by modulating peripheral adaptation. One way to evidence these changes is to measure after-effects, which has been done for orientation. We extend this approach to more complex properties such as numerosity. So far, we have described how we investigate phenomenal appearance, and sensitivity (the ability to correct identify shapes). In another set of studies, we examine the conditions under which the illusory nature of peripheral appearance affects perceptual confidence (BO2), i.e., to what extent we feel confident in what we see. In the final part (PD3) we address the plasticity mechanisms that allow the foveal visual cortex to process peripheral input after longand short-term deprivation. We do this by using Transcranial Magnetic Stimulation (TMS) in both patients with macular degeneration (MD), who lack foveal vision, and in healthy observers temporarily deprived of foveal input using a gaze-contingent display. When the foveal cortex is deprived of normal input, it may start to respond to peripheral stimuli. Furthermore, we test whether perceptual learning can alter the peripheral representation in the foveal cortex both in MD patients and in healthy observers. Overall, we propose an integrated research program, which elucidates how central and peripheral vision interact over a range of stimulus complexities (from simple 2D gratings to complex 3D shapes), representation levels (from detection to appearance and to confidence) and time frames (from immediate interactions to adaptation and to long-term plasticity). Results achived Research at the UNIBO unit focused on the broad topic of similarities and differences between central and peripheral vision, and in particular on the properties of the mechanisms that allow us to produce an integrated phenomenology of the visual world despite the anisotropies of our foveated visual system. First, we collaborated on a study investigating how the visual system integrates peripheral and foveal vision as observers move their eyes to gaze at peripheral objects (Sharvashidze et al., 2024). We used a gaze-contingent paradigm to modify the level of distortion and fuzziness of geometrical shapes as observers made saccadic eye movements toward them. Participants were asked either to judge the appearance of the stimuli before or after looking at them, or to detect the presence of trans-saccadic changes. The results showed that observers tend to perceive peripheral stimuli as relatively sharper than they actually are, indicating that peripheral appearance overcompensates for the lower sensitivity to fine details. Moreover, trans-saccadic decreases in regularity were more visible than trans-saccadic increases, suggesting that the visual system incorporates a prior for high regularity based on peripheral preview. The fact that, under certain conditions, we can have the impression of perceiving our peripheral visual field better and more extensively than warranted by our sensitivity led to the second question addressed in the project: to what extent are we aware of the illusory nature of the richness of peripheral appearance? To this end, we conducted a second study (Sharvashidze et al., 2025) investigating perceptual confidence in a scene-categorization task performed either in central or peripheral vision. In each trial, observers were exposed to a scene viewed through a central circular aperture (“window” condition) and to a scene viewed through a peripheral ring aperture (“scotoma” condition). They were required to identify the category (beach, mountain, etc.) of each scene and to indicate which of the two categorizations they were more confident in. By titrating the size of the apertures in the two conditions, we were able to evaluate confidence judgments when performance in central and peripheral vision was equalized. The results showed that, given the same performance, observers were biased toward trusting their judgments in the peripheral (“scotoma”) condition over the central (“window”) one. This indicates that, in a task such as scene categorization, which is particularly suited to global peripheral processing, observers overtrust their ability to “see” what is in their visual periphery. Thus, we showed that peripheral appearance is richer than it should be and that, at times, we are overconfident in what we see. But what is the source of this richness of peripheral vision? In a third study (Metzger et al., under review), we addressed one predictive mechanism that might contribute to enriching our peripheral experience: the memory color effect. The memory color effect is a phenomenon whereby stimuli strongly associated with a particular color (e.g., bananas with yellow, strawberries with red) tend to appear tinged with that color. This is evidenced by the fact that when observers are asked to adjust the color of a diagnostic object to gray, they tend to produce adjustments shifted in the opposite color direction relative to the associated one, in order to compensate for the endogenously driven color appearance. For example, observers will adjust the color of a banana toward a bluish hue to compensate for the illusory yellow and achieve the appearance of gray. In our study, we compared the memory color effect in central and peripheral vision and consistently found that as sensory uncertainty about object color increases in peripheral vision, so does the strength of the memory color effect. This indicates that predictive mechanisms based on stored representations contribute to colorizing the appearance of the peripheral visual field. A final study investigated similarities and differences in the mechanisms involved in attending to moving stimuli in central vision, via smooth pursuit eye movements, and in peripheral vision, via covert attentional tracking (Songlin et al., 2025). In this task, participants fixated on a point on the screen while attending to a stimulus moving in the periphery at different eccentricities and speeds, with EEG signals recorded throughout. The results indicated that alpha power was reduced as the stimulus moved further into the periphery, and that alpha suppression was larger contralateral to the stimulus location. A comparison with the existing literature on pursuit tracking suggests that attention plays a key role in both peripheral tracking and foveal pursuit, although some differences do emerge. For instance, target speed has a larger effect on the attentional requirements of pursuit, possibly because tracking imprecision leads to fixation offsets and catch-up saccades. Overall, the work conducted by the UNIBO unit contributed to shedding light on the similarities and differences between central and peripheral vision, showing that predictive mechanisms enrich the appearance of the peripheral world and that, under some conditions, we are not aware that this is the case. While we elucidated one specific predictive mechanism, color prediction based on object semantics, many others likely contribute, such as trans-saccadic and other forms of learning, which deserve further investigation. Publications: - Sharvashidze, N., Valsecchi, M., & Schütz, A. C. (2024). Transsaccadic perception of changes in object regularity. Journal of Vision, 24(13), 3-3. - Sharvashidze, N., Toscani, M., & Valsecchi, M. (2025). Peripheral overconfidence in a scene categorization task. Journal of vision, 25(10), 2-2. - Songlin, Q., Xia, X., Chen, J., & Valsecchi, M. (2025). Attentional tracking reduces cortical alpha oscillations. Scientific Reports, 15(1), 34067. - Metzger, A., Valsecchi, M., & Toscani, M. (under review). Enhanced memory colour in peripheral vision: a possible compensation for chromatic loss.
Dettagli del progetto
Responsabile scientifico: Matteo Valsecchi
Strutture Unibo coinvolte:
Dipartimento di Psicologia "Renzo Canestrari"
Coordinatore:
Università degli Studi di PADOVA(Italy)
Contributo totale Unibo: Euro (EUR) 78.174,00
Durata del progetto in mesi: 24
Data di inizio
05/10/2023
Data di fine:
28/02/2026
