Anisotropy in MT for disparity motion segmentation 1 Population anisotropy in area MT explains a perceptual difference 1 between near and far disparity motion segmentation

22 Segmentation of the visual scene into relevant object components is a 23 fundamental process for successfully interacting with our surroundings. Many visual 24 cues, including motion and binocular disparity, support segmentation, yet the 25 mechanisms utilizing these cues are unclear. We used a psychophysical motion 26 discrimination task in which noise dots were displaced in depth to investigate the role of 27 segmentation through disparity cues in visual motion stimuli (Experiment 1). We found a 28 subtle, but significant, bias indicating that near disparity noise disrupted the segmentation 29 of motion more than equidistant far disparity noise. A control experiment showed that 30 the near-far difference could not be attributed to attention (Experiment 2). To account for 31 the near-far bias, we constructed a biologically constrained model using recordings from 32 neurons in the middle temporal area (MT) to simulate human observers’ performance on 33 Experiment 1. Performance of the model of MT neurons showed a near-disparity skew 34 similar to that shown by human observers. To isolate the cause of the skew, we simulated 35 performance of a model containing units derived from properties of MT neurons, using 36 phase-modulated Gabor disparity tuning. Using a skewed-normal population distribution 37 of preferred disparities, the model reproduced the elevated motion discrimination 38 thresholds for near-disparity noise, whereas a skewed-normal population of phases 39 (creating individually asymmetric units) did not lead to any performance skew. Results 40 from the model suggest that the properties of neurons in area MT are computationally 41 sufficient to perform disparity segmentation during motion processing, and produce 42 similar disparity biases as those produced by human observers. 43