Abstract
Processing effectiveness trough time in visual object recognition was investigated using random temporal stimulus sampling. Twelve young adults named photographs of familiar objects from the Bank of Standardized Stimuli. Stimuli were made by a linear combination of the target image and high density white visual noise. Signal-to-noise ratio varied throughout the 200 ms stimulus duration. A new temporal sampling function was generated on each trial by the integration of random amplitude and phase sinusoidal waves of frequency between 5 and 60 Hz. Individual temporal classification images were calculated by subtracting the weighted sum of the signal/noise ratio associated to errors from that associated to correct responses. Time-frequency classification images were obtained by applying the same procedure on the outcome of time-frequency analyses applied to the sampling functions of each trial. The mean temporal classification image indicates that processing effectiveness is lowest at target onset, then rises to a peak at 79–88 ms to then gradually decline to about its initial level until 200 ms. This temporal pattern was highly variable across participants. The highest effectiveness peak revealed by the mean time-frequency classification image occurred in the 0–67 ms time window for 25–35 Hz stimulus oscillations. The amplitude of this peak slowly declined from 67 to 167 ms while gradually shifting towards lower frequencies down to 10–15 Hz. Stimulus oscillation frequencies around 35–55 Hz in the 67–167 ms range were associated with the lowest processing effectiveness. However, a subset of these frequencies (40–50 Hz) led to high effectiveness between 133–200 ms. This time-frequency pattern was remarkably similar across participants, with a mean between-subjects correlation of .93. The present results indicate rapid variations of visual encoding effectiveness in the initial 200 ms of stimulus exposure and suggests that the time-frequency classification images tap a highly fundamental aspect of visual processing.
Acknowledgement: Centre de recherche en science naturelle et en génie du Canada