Perceptual Learning at a Conceptual Level

Wang, Rui; Wang, Jie; Zhang, Jun-Yun; Xie, Xinyu; Yang, Yuxiang; Luo, Shuhan; Yu, Cong; Li, Wu

doi:10.1523/jneurosci.2732-15.2016

Cited by 79 publications

(73 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In addition, we also observed specificities to the trained direction, speed, stimulus size, and contrast. These results are in line with the previous findings that VPL is generally vulnerable to the variations in basic feature dimensions and argue against plasticity in high-level brain areas that represent non-sensory cognitive factors, such as general task statistics and decision rules 15,16,17 .…”

Section: Discussionsupporting

confidence: 92%

Disentangling locus of perceptual learning in the visual hierarchy of motion processing

Zhang

Tadin

2018

Preprint

View full text Add to dashboard Cite

32 Visual perceptual learning (VPL) can lead to long-lasting perceptual improvements. 33 While the efficacy of VPL is well established, there is still a considerable debate about 34 what mechanisms underlie the effects of VPL. Much of this debate concentrates on where 35 along the visual processing hierarchy behaviorally relevant plasticity takes place. Here, 36we aimed to tackle this question in context of motion processing, a domain where links 37 between behavior and processing hierarchy are well established. Specifically, we took 38 advantage of an established transition from component-dependent representations at the 39 earliest level to pattern-dependent representations at the middle-level of cortical motion 40 processing. We trained two groups of participants on the same motion direction 41 identification task using either grating or plaid stimuli. A set of pre-and post-training 42 tests was used to determine the degree of learning specificity and generalizability. This 43 approach allowed us to disentangle contributions from both low-and mid-level motion 44 processing, as well as high-level cognitive changes. We observed a complete bi-45 directional transfer of learning between component and pattern stimuli as long as they 46 shared the same apparent motion direction. This result indicates learning-induced 47

show abstract

Section: Discussionsupporting

confidence: 92%

Disentangling locus of perceptual learning in the visual hierarchy of motion processing

Zhang

Tadin

2018

Preprint

View full text Add to dashboard Cite

show abstract

“…Although beyond the scope of the present study, future modeling targets can be considered in pursuit of many of perceptual learning phenomena. For example, DNNs may be used to replicate other psychophysical phenomena, including disruption (Seitz et al, 2005), roving (Zhang et al, 2008;Tartaglia et al, 2009;Hussain et al, 2012), double training Zhang et al, 2010), and the effects of attention (Ahissar and Hochstein, 1993;Byers and Serences, 2012;Bays et al, 2015;Donovan et al, 2015) and adaptation (Harris et al, 2012). Moreover, small variations in training procedures can lead to dramatic changes in learning outcome (Hung and Seitz, 2014); therefore, it is important for future simulations to take into consideration how such details may affect learning in DNNs.…”

Section: Discussionmentioning

confidence: 99%

Deep Neural Networks for Modeling Visual Perceptual Learning

Wenliang

Seitz

2018

J. Neurosci.

View full text Add to dashboard Cite

Understanding visual perceptual learning (VPL) has become increasingly more challenging as new phenomena are discovered with novel stimuli and training paradigms. Although existing models aid our knowledge of critical aspects of VPL, the connections shown by these models between behavioral learning and plasticity across different brain areas are typically superficial. Most models explain VPL as readout from simple perceptual representations to decision areas and are not easily adaptable to explain new findings. Here, we show that a well -known instance of deep neural network (DNN), whereas not designed specifically for VPL, provides a computational model of VPL with enough complexity to be studied at many levels of analyses. After learning a Gabor orientation discrimination task, the DNN model reproduced key behavioral results, including increasing specificity with higher task precision, and also suggested that learning precise discriminations could transfer asymmetrically to coarse discriminations when the stimulus conditions varied. Consistent with the behavioral findings, the distribution of plasticity moved toward lower layers when task precision increased and this distribution was also modulated by tasks with different stimulus types. Furthermore, learning in the network units demonstrated close resemblance to extant electrophysiological recordings in monkey visual areas. Altogether, the DNN fulfilled predictions of existing theories regarding specificity and plasticity and reproduced findings of tuning changes in neurons of the primate visual areas. Although the comparisons were mostly qualitative, the DNN provides a new method of studying VPL, can serve as a test bed for theories, and assists in generating predictions for physiological investigations. Visual perceptual learning (VPL) has been found to cause changes at multiple stages of the visual hierarchy. We found that training a deep neural network (DNN) on an orientation discrimination task produced behavioral and physiological patterns similar to those found in human and monkey experiments. Unlike existing VPL models, the DNN was pre-trained on natural images to reach high performance in object recognition, but was not designed specifically for VPL; however, it fulfilled predictions of existing theories regarding specificity and plasticity and reproduced findings of tuning changes in neurons of the primate visual areas. When used with care, this unbiased and deep-hierarchical model can provide new ways of studying VPL from behavior to physiology.

show abstract

“…Perceptual learning is typically obtained in an experimental context by providing feedback on discrimination performance, on a trial-by-trial basis1. Several mechanisms underlying perceptual learning have been proposed, from refined encoding of sensory information to improved decision making2345. Interestingly, learning is not necessarily confined to a specific stimulus feature: learning to discriminate one of the features that defines a sensory stimulus (e.g., the orientation of a visual stimulus) sometimes transfers, or generalizes, to a different feature of the same stimulus (e.g., the contrast of the same visual stimulus)67.…”

mentioning

confidence: 99%

Perceptual learning to discriminate the intensity and spatial location of nociceptive stimuli

Mancini

Dolgevica

Steckelmacher

et al. 2016

Sci Rep

View full text Add to dashboard Cite

Accurate discrimination of the intensity and spatial location of nociceptive stimuli is essential to guide appropriate behaviour. The ability to discriminate the attributes of sensory stimuli is continuously refined by practice, even throughout adulthood - a phenomenon called perceptual learning. In the visual domain, perceptual learning to discriminate one of the features that define a visual stimulus (e.g., its orientation) can transfer to a different feature of the same stimulus (e.g., its contrast). Here, we performed two experiments on 48 volunteers to characterize perceptual learning in nociception, which has been rarely studied. We investigated whether learning to discriminate either the intensity or the location of nociceptive stimuli (1) occurs during practice and is subsequently maintained, (2) requires feedback on performance, and (3) transfers to the other, unpractised stimulus feature. First, we found clear evidence that perceptual learning in discriminating both the intensity and the location of nociceptive stimuli occurs, and is maintained for at least 3 hours after practice. Second, learning occurs only when feedback is provided during practice. Finally, learning is largely confined to the feature for which feedback was provided. We discuss these effects in a predictive coding framework, and consider implications for future studies.

show abstract

Perceptual Learning at a Conceptual Level

Cited by 79 publications

References 51 publications

Disentangling locus of perceptual learning in the visual hierarchy of motion processing

Disentangling locus of perceptual learning in the visual hierarchy of motion processing

Deep Neural Networks for Modeling Visual Perceptual Learning

Perceptual learning to discriminate the intensity and spatial location of nociceptive stimuli

Contact Info

Product

Resources

About