Glass patterns are texture stimuli made by pairing randomly placed dots with partners at specific offsets. The strong percept of global form that arises from sparse local orientation cues has made these patterns the subject of psychophysical investigations, yet neuronal responses to Glass patterns have not been studied.
We measured the responses of neurons in macaque V1 and V2 to dynamic translational Glass patterns. Response and selectivity to these patterns was predictable from that to gratings. Responses to translational, concentric and radial patterns extending outside the classical receptive field (CRF) were very similar to those elicited by a comparable translational pattern within the CRF.
We computed the expected responses for a receptive field model to translational Glass patterns, and found that the complexity of V1 tuning curves could be understood in terms of the responses of linear filters to dot pairs. This modeling connects our understanding of V1 receptive fields as rectified, quasi-linear filters with results from psychophysical studies of Glass patterns.
Responses to plaid stimuli, created by superimposing two gratings, have often been studied in visual cortex. In V1, responses to a grating of preferred orientation can be suppressed by superimposing an orthogonal mask or by a parallel stimulus outside the CRF. We used a random, dynamic stimulus to compare the timing of these effects. Suppression from within the CRF had dynamics which were much faster than suppression originating from the surround. This points to different mechanisms and constrains models of their circuitry.
In macaque MT, neurons can be classified as pattern or component direction selective (PDS or CDS) if they are selective for the net motion of a plaid or its oriented components, respectively. We explored MT response dynamics using a rapidly-changing pseudo-random sequence of gratings and plaids. CDS neurons tend to show faster dynamics than PDS neurons, causing the population motion response to evolve after motion onset. The initial response is predominantly driven by component motion, and after some tens of milliseconds a reliable pattern motion response emerges. This suggests that the neuronal circuit for a PDS neuron involves more computation than for a CDS neuron.