Transfer in Rule-Based Category Learning Depends on the Training Task

Statistical Learning in Vision

Vision and learning have long been considered to be two areas of research linked only distantly. However, recent developments in vision research have changed the conceptual definition of vision from a signal-evaluating process to a goal-oriented interpreting process, and this shift binds learning, together with the resulting internal representations, intimately to vision. In this review, we consider various types of learning (perceptual, statistical, and rule/abstract) associated with vision in the past decades and argue that they represent differently specialized versions of the fundamental learning process, which must be captured in its entirety when applied to complex visual processes. We show why the generalized version of statistical learning can provide the appropriate setup for such a unified treatment of learning in vision, what computational framework best accommodates this kind of statistical learning, and what plausible neural scheme could feasibly implement this framework. Finally, we list the challenges that the field of statistical learning faces in fulfilling the promise of being the right vehicle for advancing our understanding of vision in its entirety. Expected final online publication date for the Annual Review of Vision Science, Volume 8 is September 2022. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

Learning and generalization of within-category representations in a rule-based category structure

The task requirements during the course of category learning are critical for promoting within-category representations (e.g., correlational structure of the categories). Recent data suggest that for unidimensional rule-based structures, only inference training promotes the learning of within-category representations, and generalization across tasks is limited. It is unclear if this is a general feature of rule-based structures, or a limitation of unidimensional rule-based structures. The present work reports the results of three experiments further investigating this issue using an exclusive-or rule-based structure where successful performance depends upon attending to two stimulus dimensions. Participants were trained using classification or inference and were tested using inference. For both the classification and inference training conditions, within-category representations were learned and could be generalized at test (i.e., from classification to inference) and this result was dependent upon a congruence between local and global regions of the stimulus space. These data further support the idea that the task requirements during learning (i.e., a need to attend to multiple stimulus dimensions) are critical determinants of the category representations that are learned and the utility of these representations for supporting generalization in novel situations.

Positive, negative, or all relative? Evaluative conditioning of ambivalence

In evaluative conditioning (EC), the pairing of a positively or negatively valenced stimulus (US) with a neutral stimulus (CS) leads to a corresponding change in liking of the CS. EC research so far has concentrated on using unambiguously positive or negative USs. However, attitude objects are often ambivalent, i.e., can simultaneously possess positive and negative features. The present research investigated whether ambivalence can be evaluatively conditioned and whether contingency awareness moderates this effect. In two studies, positive, negative, neutral, and ambivalent USs were paired with affectively neutral CSs. Results showed standard EC effects that were moderated by contingency awareness. Most interestingly, EC effects were also obtained for the ambivalent USs, indicating that ambivalence can indeed be conditioned. However, contingency awareness seemed to play a lesser role in ambivalence conditioning. Ambivalence EC effects were obtained on subjective and objective direct measures of ambivalence as well as on a more indirect measure.

Enhancement of motor skill learning by a combination of ideal model-observation and self-observation

[Purpose] In sports physical therapy, video of a patient’s movement or of a skilled model’s movement has been used as observational learning methods for injury prevention and movement modification. Positive effects of model video observation have been reported. This study aimed to clarify the effect on motor skill learning using a combination of model-observation and self-observation, which is thought to act as an enhanced method for active error detection by comparing model-observation and self-observation alone for acquisition of correct sports movement. [Subjects and Methods] Forty-five healthy females were randomly allocated into three groups comprising model- and self-observation, model-observation, and self-observation. The motor task performed was a half golf swing using an elastic club. Shoulder grip angle between both shoulder lines and the acromia grip strength were measured as an index of body rotation using a three-dimensional motion analyzer. Change in the shoulder grip angle in the three groups was analyzed at pre-, immediate delayed retention, and delayed retention tests. [Results] A significant difference in shoulder grip angle was observed among the three groups for the immediate delayed retention test. The combined model and self-observation group had a value closer to 90 degrees compared to the other two groups. [Conclusion] Observation combining model and self-observation exerted a positive effect on short-term motor skill learning.