Enabling complete transfer of perceptual learning across orientations in foveal vision through double training

Low-level sensory plasticity during task-irrelevant perceptual learning: Evidence from conventional and double training procedures

“…Comparison of the current results with those from task-relevant paradigms, namely Xiao et al (2008) and Zhang et al (2009), points out that the presence of attention may be an important factor in the recruitment of higher-level central brain sites for perceptual learning. The fact that task-irrelevant learning bypasses some effects of the attentional system may explain why double training has different impact on task-irrelevant than for task-relevant perceptual learning.…”

“…While in Experiment 1 we found that task-irrelevant learning of direction discrimination is specific to the direction and contrast polarity of unattended motion stimuli spatiotemporally paired with targets of an RSVP task, in light of the task-relevant double training results (Xiao et al, 2008; Zhang et al, 2009), it is possible that our observed lack of transfer to the opposite polarity could just simply be due to the lack of priming that polarity during the RSVP sessions. To address this possibility, which has not previously been tested in the context of TIPL, we performed Experiment 2.…”

“…Some recent studies have brought into question the view that featural specificity is a good indicator of early sensory plasticity (Xiao et al, 2008; Zhang et al, 2009; Dosher & Lu 1998) and explored alternative hypotheses to explain featural specificity. Any perceptual task involves the processing of a perceptual stimulus, which can implicitly encode many individual features such as contrast, retinal location, and orientation.…”

“…The results of this “double training” are difficult to explain in terms of plasticity in retinotopic early visual areas as the basis for featurally-specific perceptual learning. Further, Zhang et al (2009) showed that the unrelated task in the second training could be very coarse identification of a stimulus comprising the transfer feature, such as discriminating if the presented stimulus is a Gabor exposed at the transfer orientation or the letter C, and concluded that stimulus exposure was a sufficient form of double training to produce transfer of learning.…”

“…While some studies have argued that perceptual learning can take place at early stages of sensory processing (Furmanski, Schluppeck, & Engel, 2004; Pourtois et al, 2008; Schoups et al, 2001; Watanabe et al, 2002), other studies have proposed that most of the learning takes place in the readout from sensory to decision stages (Dosher & Lu, 1998; Law & Gold, 2008; Smirnakis et al, 2005) or in higher non-retinotopic, central stages (Xiao et al, 2008; Zhang et al, 2009). A likely resolution of this long-standing controversy is that the degree to which learning occurs at various processing stages may depend on details of the procedure, such as training task difficulty (Ahissar & Hochstein, 1997; Liu & Weinshall, 2000), transfer task precision (Jeter et al, 2009), and training procedure (Xiao et al, 2008; Zhang et al, 2009). …”