Perceptual Learning

Perceptual learning is the process by which our neural representations of perceptual patterns, object, scenes and events change as a fuction of experience. In the VCNLab, we have investigated perceptual learning in a range of contexts and participants. 

For example, our initial paper on the topic, Gold, Bennett & Sekuler (1999, Nature) was the first to show that perceptual learning for faces and visual textures occurs as a result of increasing the efficiency of processing, not by reducing levels of internal noise. More recently, we showed that one can use the response classification technique to visualize how the behavioural receptive fields change as a function of learning. Both before and after learning faces, observers rely on information about the eye and brow region, but after learning, observers weight information more ideally. 

 

This pattern of results has an interesting parallel with the effects of inversion on face perception, consistent with the idea that upright faces are easier to recognize than inverted faces because we have had more experience with upright faces. Our paper on learning with textures and houses supports that idea -- showing that we can develop inversion effects as a function of learning, and that the size of the inversion effect is not necessarily a good indicator of configural processing. 

We also have investigated the effects of learning on dividing attention in younger and older participants. Before training, older observers have a more difficult time dividing attention (multitasking) than younger observers. After training, older observers learn to multitask -- performing two visual tasks simultaneously as well as they can perform one at a time.

 

On-going studies of perceptual learning in the VCNLab are investigating the way factors such as sleep, stimulus consistent, and numbers of trials affect perceptual learning; how perceptual learning and attention interact; and how perceptual learning changes the spatial (fMRI) and temporal (EEG) patterns of processing in the brain.

 

Recent sample publications.

Pachai, M. V., Bennett, P. J., & Sekuler, A. B. (2018). The effect of training with inverted faces on the selective use of horizontal structure. Vision research.

Serrano, A., Hashemi, A., Sekuler, A., & Bennett, P. (2017). Ruling out task difficulty in the context-generalization of texture perceptual learning. Journal of Vision, 17(10), 504-504.

Hashemi, A., Sekuler, A., & Bennett, P. (2017). Extensive training of orientation filtered textures increases generalization of learning. Journal of Vision, 17(10), 509-509.

Hashemi, A., Pachai, M., Sekuler, A., & Bennett, P. (2016). Learning to generalize stimulus-specific learning across contexts. Journal of Vision, 16(12), 1110-1110.

JS Husk, PJ Bennett & AB Sekuler (2007) Houses and textures: Investigating the characteristics of inversion effects.  Vision Research , 47, 3350-3359.

 

VCNLab research on Perceptual Learning is funded by NSERC, CIHR, and the Canada Research Chair program.