Perceptual Learning: Complete Transfer across Retinal Locations

Temporal perceptual learning shows specificity to the visual or auditory modality, or asymmetric partial transfer from auditory to visual, but not vice versa. These findings are interpreted as evidence for distributed, rather than central, temporal processing (Ivry & Schlerf, 2008). For visual perceptual learning, location and orientation specificity can be eliminated with double training, indicating that learning specificity may not be used to infer the mechanisms of perceptual learning (e.g., Xiao et al., 2008). Here we investigated whether double training can eliminate the modality specificity in temporal learning. We first replicated asymmetric partial transfer results between auditory and visual learning of a temporal-interval discrimination (TID) task with standard training. The standard interval was marked by a pair of auditory beeps or visual gratings at 100-ms. The subjects practiced either the auditory or the visual TID task for 5 sessions. Visual TID learning had no impact on auditory TID performance (p=0.65), while auditory TID learning improved visual TID performance (p=0.005), although not as much as direction visual TID learning (p=0.028). However, complete learning transfer was evident with double-training. When visual TID learning was paired with an auditory frequency discrimination task at the same 100-ms interval, auditory TID performance was improved similarly to direct auditory training (p=0.051), indicating complete cross-modal learning transfer. Similarly, when auditory TID learning was paired with a visual contrast discrimination task at the same 100-ms interval, visual TID performance was improved equally as direct visual training (p=0.95), again indicating complete cross-modal learning transfer. In both cases we found no significant impact of practicing auditory frequency discrimination or visual contrast discrimination alone on TID performance. Our results suggest mutual and nearly complete learning transfer of TID learning between visual and auditory modalities, which are consistent with a central temporal processing mechanism shared by different modalities.

Perceptual learning is usually feature-specific. Recently, we showed that perceptual learning is even specific for the motor response type. In a three-line bisection task, participants indicated whether the central line was offset either to the left or right by pressing a left or a right button, respectively. We found no transfer when the same participants adjusted the offset by using a computer mouse. Here, we first show that perceptual learning with mouse adjustments transfers to the untrained hand, but only for the trained adjustment condition. There was no transfer to the button press conditions, neither for the trained nor the untrained hand. Second, we show that a double training procedure enables transfer from the mouse adjustment to the button press condition. Hence, the specificity of perceptual learning to the motor response type can be overcome by double training as it is the case for visual features. Our results suggest that during perceptual learning, perceptuo-decisional signals are encoded together with the corresponding actions.

Subsecond time perception has been frequently attributed to modality-specific timing mechanisms that would predict no cross-modal transfer of temporal perceptual learning. In fact, perceptual learning of temporal interval discrimination (TID) reportedly shows either no cross-modal transfer, or asymmetric transfer from audition to vision, but not vice versa. However, here we demonstrate complete cross-modal transfer of auditory and visual TID learning using a double training paradigm. Specifically, visual TID learning transfers to and optimizes auditory TID when the participants also receive exposure to the auditory temporal interval by practicing a functionally orthogonal near-threshold tone frequency discrimination task at the same trained interval. Auditory TID learning also transfers to and optimizes visual TID with additional practice of an orthogonal near-threshold visual contrast discrimination task at the same trained interval. Practicing these functionally orthogonal tasks per se has no impact on TID thresholds. We interpret the transfer results as indications of a supramodal representation of subsecond time. Moreover, because TID learning shows complete transfer between modalities with vastly different temporal precisions, the sub-second time presentation must be conceptual. Double training may refine this supramodal and conceptual subsecond time representation and connect it to a new sense to improve time perception.

In this review, we explore how reward signals shape perceptual learning in animals and humans. Perceptual learning is the well-established phenomenon by which extensive practice elicits selective improvement in one's perceptual discrimination of basic visual features, such as oriented lines or moving stimuli. While perceptual learning has long been thought to rely on 'top-down' processes, such as attention and decision-making, a wave of recent findings suggests that these higher-level processes are, in fact, not necessary. Rather, these recent findings indicate that reward signals alone, in the absence of the contribution of higher-level cognitive processes, are sufficient to drive the benefits of perceptual learning. Here, we will review the literature tying reward signals to perceptual learning. Based on these findings, we propose dual underlying mechanisms that give rise to perceptual learning: onemechanism that operates 'automatically' and is tied directly to reward signals, and another mechanism that involves more'top-down', goal-directedcomputations.

The brain site of perceptual learning has been constantly debated. Recent evidence showing complete learning transfer to an untrained retinal location or orthogonal orientation suggests that perceptual learning occurs in high-level brain areas (Xiao et al., 2008; Zhang et al., 2010). Contradictorily, ERP C1 changes after learning are used as evidence for V1 plasticity because of C1’s early onset (Pourtois et al., 2008; Bao et al., 2010). However, C1 can be top-down modulated (Rauss et al., 2011), which is especially likely after an observer practices/learns the same stimuli/task for several sessions. Here we measured C1 changes associated with learning transfer at a completely untrained location, with the assumption that any C1 changes observed here would indicate top-down modulation, rather than V1 plasticity. Twelve subjects were trained to discriminate grating orientation (ref_ori=36°) for five days. Two locations at 5-deg eccentricity in diagonal quadrants where obvious C1 components were evoked were used as training and transfer locations, respectively. EEGs responding to suprathreshold orientation judgment (36° or 36°±15°) were acquired pre- and post-training in separate sessions. Training improved orientation thresholds significantly (46.3%, p<0.001). Performance at the untrained location (using pretest threshold at the trained location as baseline) was also significantly improved, although by a less amount (28.3%, p<0.001). Importantly, ERP C1 increased significantly after training at not only the trained location (p=0.009), but also the untrained location (p=0.001)! These results support our hypothesis that ERP C1 is top-down modulated in perceptual learning, and that C1 changes do not necessarily indicate V1 plasticity. Bao et al. (2010) and Pourtois et al. (2008) also show no significant C1 changes if learning does not transfer to the untrained location. These associations between C1 changes and learning-induced performance improvement together suggest that C1 is likely top-down modulated by high-level perceptual learning. Meeting abstract presented at VSS 2013

How and under what circumstances the training effects of perceptual learning (PL) transfer to novel situations is critical to our understanding of generalization and abstraction in learning. Although PL is generally believed to be highly specific to the trained stimulus, a series of psychophysical studies have recently shown that training effects can transfer to untrained conditions under certain experimental protocols. In this article, we present a brain-inspired, neuromorphic computational model of the Where-What visuomotor pathways which successfully explains both the specificity and transfer of perceptual learning. The major architectural novelty is that each feature neuron has both sensory and motor inputs. The network of neurons is autonomously developed from experience, using a refined Hebbian-learning rule and lateral competition, which altogether result in neuronal recruitment. Our hypothesis is that certain paradigms of experiments trigger two-way (descending and ascending) off-task processes about the untrained condition which lead to recruitment of more neurons in lower feature representation areas as well as higher concept representation areas for the untrained condition, hence the transfer. We put forward a novel proposition that gated self-organization of the connections during the off-task processes accounts for the observed transfer effects. Simulation results showed transfer of learning across retinal locations in a Vernier discrimination task in a double-training procedure, comparable to previous psychophysical data (Xiao et al., 2008). To the best of our knowledge, this model is the first neurally-plausible model to explain both transfer and specificity in a PL setting.

Learning to better discriminate a specific visual feature (i.e., a specific orientation in a specific region of space) has been associated with plasticity in early visual areas (sensory modulation) and with improvements in the transmission of sensory information from early visual areas to downstream sensorimotor and decision regions (enhanced readout). However, in many real-world scenarios that require perceptual expertise, observers need to efficiently process numerous exemplars from a broad stimulus class as opposed to just a single stimulus feature. Some previous data suggest that perceptual learning leads to highly specific neural modulations that support the discrimination of specific trained features. However, the extent to which perceptual learning acts to improve the discriminability of a broad class of stimuli via the modulation of sensory responses in human visual cortex remains largely unknown. Here, we used functional MRI and a multivariate analysis method to reconstruct orientation-selective response profiles based on activation patterns in the early visual cortex before and after subjects learned to discriminate small offsets in a set of grating stimuli that were rendered in one of nine possible orientations. Behavioral performance improved across 10 training sessions, and there was a training-related increase in the amplitude of orientation-selective response profiles in V1, V2, and V3 when orientation was task relevant compared with when it was task irrelevant. These results suggest that generalized perceptual learning can lead to modified responses in the early visual cortex in a manner that is suitable for supporting improved discriminability of stimuli drawn from a large set of exemplars.

Visual perceptual learning (VPL) is defined as long-term improvement on a visual task as a result of visual experience (Sasaki et al., 2012). The improvement is usually highly specific to the location where the trained feature is presented, which refers to location specificity. Location specificity is regarded as manifestation of changes in early visual areas in association with VPL. Recently, Xiao et al. (2008) have demonstrated that location specificity of VPL is abolished by double training. For example, location-specific contrast discrimination learning can be rendered completely transferrable from the upper left to a new location in the lower-right quadrant if the new location is trained with an additional orientation discrimination task. This evidence challenges traditional views towards location specificity. However, the mechanism underlying double training remains unclear. In the current study, we applied a double training paradigm to investigate why it occurs. Subjects performed orientation discrimination training at the upper left quadrant for 6 days and then performed contrast discrimination training at the lower right quadrant for another 6 days. We found that, after the initial orientation discrimination training, the threshold significantly decreased at the trained location. In addition, the performance enhancement partially transferred to the lower right quadrant. However, hardly any transfer occurred at the upper right quadrant. The additional training led to some amount of improvement in orientation discrimination at the newly trained location, but not in the upper right quadrant. The result suggests that it is easier to trigger the transfer of orientation discrimination to the diagonal quadrant compared with the parallel quadrant. The effect of double training could be at least partially confounded by the fact that VPL is more likely to be transferred in diagonal direction than in parallel direction. Meeting abstract presented at VSS 2014

We investigated intermanual transfer and long-term retention of practice-related perceptual learning in the domain of tactile hyperacuity. Subjects discriminated a row of three dots in which the central dot was offset laterally from a row without such offset. Performance at the right index fingerpad improved with practice. Practice effects transferred essentially completely to the left index fingerpad. When tested some months later at the right index fingerpad, long-term retention of learning was limited and further practice was required to stabilize discrimination thresholds. Intermanual transfer of tactile learning appears to be a general phenomenon, while long-term retention appears to be limited in hyperacuity tasks.

People tend to associate abstract visual features with basic taste qualities. This narrative historical review critically evaluates the literature on these associations, often referred to as crossmodal correspondences, between basic tastes and visual design features such as color hue and shape curvilinearity. The patterns, discrepancies, and evolution in the development of the research are highlighted while the mappings that have been reported to date are summarized. The review also reflects on issues of cross-cultural validity and deviations in the matching patterns that are observed when correspondences are assessed with actual tastants versus with verbal stimuli. The various theories that have been proposed to account for different classes of crossmodal correspondence are discussed, among which the statistical and affective (or emotional-mediation) accounts currently appear most promising. Several critical research questions for the future are presented to address the gaps that have been identified in the literature and help validate the popular theories on the origin and operations of visual-taste correspondences.

Word length effects in Hebrew

1. Human perceptual learning in discrimination of the oblique orientation was studied using psychophysical methods. Subjects were trained daily to improve their ability to identify the orientation of a circular 2.5 deg diameter unidimensional noise field. Dramatic improvements in sensitivity to contour orientation occurred over a period of 15-20 days. The improved performance persisted for several months. Improvement was more evident between daily sessions than within sessions. This was partly due to fatigue interfering with the learning effect. Moreover, a consolidation period seemed to be required. 2. Improvement was restricted to the position of the stimulus being trained. This position dependency of the learning effect proved very precise. After training at a specific stimulus position, merely displacing the stimulus to an adjacent position caused a marked increase in thresholds. 3. No transfer of the training effect was observed between orientations. Following a shift of 90 deg away from the trained orientation, performance fell, even below the initial level. 4. We observed complete to almost complete transfer between the two eyes. 5. Our results suggest plastic changes at a level of the visual processing stream where input from both eyes has come together, but where generalization for spatial localization and orientation has not yet occurred.

Perceptual learning refers to an improvement on a perceptual task after repeated exposure to a stimulus. It has been shown that attention can play an important role in perceptual learning. Recently, it has been suggested that training can lead to increased suppression of information that is continuously irrelevant, and that this attention-based suppression plays an important role in more efficient noise exclusion. Here we investigate this claim. Observers performed a visual speed-discrimination task for 5 consecutive days. After training, sensitivity to motion directions that were relevant, irrelevant, or neutral toward the training task was assessed by measuring motion coherence thresholds. In addition, perceptual dominance during binocular rivalry was assessed for combinations of the three motion directions. The results showed that sensitivity to the task-relevant feature increased due to training. That is, motion coherence thresholds were selectively lowered for the task-relevant feature. Interestingly, the feature that was task-irrelevant during training was more strongly suppressed during binocular rivalry: The mean perceptual dominance of this feature was selectively decreased. Our results show that task-irrelevant information that potentially interferes with the primary task during learning gets more strongly suppressed. Furthermore, our results add new evidence in support of the claim that mechanisms involved in visual attention and binocular rivalry overlap.

Patients with age-related macular degeneration (AMD) or hereditary macular dystrophies (JMD) rely on an efficient use of their peripheral visual field. We trained eight AMD and five JMD patients to perform a texture-discrimination task (TDT) at their preferred retinal locus (PRL) used for fixation. Six training sessions of approximately one hour duration were conducted over a period of approximately 3 weeks. Before, during and after training twelve patients and twelve age-matched controls (the data from two controls had to be discarded later) took part in three functional magnetic resonance imaging (fMRI) sessions to assess training-related changes in the BOLD response in early visual cortex. Patients benefited from the training measurements as indexed by significant decrease (p = 0.001) in the stimulus onset asynchrony (SOA) between the presentation of the texture target on background and the visual mask, and in a significant location specific effect of the PRL with respect to hit rate (p = 0.014). The following trends were observed: (i) improvement in Vernier acuity for an eccentric line-bisection task; (ii) positive correlation between the development of BOLD signals in early visual cortex and initial fixation stability (r = 0.531); (iii) positive correlation between the increase in task performance and initial fixation stability (r = 0.730). The first two trends were non-significant, whereas the third trend was significant at p = 0.014, Bonferroni corrected. Consequently, our exploratory study suggests that training on the TDT can enhance eccentric vision in patients with central vision loss. This enhancement is accompanied by a modest alteration in the BOLD response in early visual cortex.

Following the onset of central vision loss, most patients develop an eccentric retinal location outside the affected macular region, the preferred retinal locus (PRL), as their new reference for visual tasks. The first goal of this article is to present behavioral evidence showing the presence of experience-dependent plasticity in people with central vision loss. The evidence includes the presence of oculomotor re-referencing of fixational saccades to the PRL; the characteristics of the shape of the crowding zone (spatial region within which the presence of other objects affects the recognition of a target) at the PRL are more "foveal-like" instead of resembling those of the normal periphery; and the change in the shape of the crowding zone at a para-PRL location that includes a component referenced to the PRL. These findings suggest that there is a shift in the referencing locus of the oculomotor and the sensory visual system from the fovea to the PRL for people with central vision loss, implying that the visual system for these individuals is still plastic and can be modified through experiences. The second goal of the article is to demonstrate the feasibility of applying perceptual learning, which capitalizes on the presence of plasticity, as a tool to improve functional vision for people with central vision loss. Our finding that visual function could improve with perceptual learning presents an exciting possibility for the development of an alternative rehabilitative strategy for people with central vision loss.