Human, horizontal, smooth-pursuit eye movements were recorded by the search coil method in response to Rashbass step-ramp stimuli of 5 to 30 deg/s. Eye velocity records were analyzed by measuring features such as the time, velocity and acceleration of the point of peak acceleration, the time and velocity of the peaks and troughs of ringing and steady-state velocity. These values were averaged and mean responses reconstructed. Three normal subjects were studied and their responses averaged. All showed a peak acceleration-velocity saturation. All had ringing frequencies near 3.8 Hz and the mean steady-state gain was 0.95. It is argued that a single, linear forward path with any transfer function G(s) and a 100 ms delay (latency) cannot simultaneously simulate the initial rise of acceleration and ring at 3.8 Hz based on a Bode analysis. Also such a simple negative feedback model cannot have a steady-state gain greater than 1.0; a situation that occurs frequently experimentally. L.R. Young's model, which employs internal positive feedback to eliminate the built-in unity negative feedback, was felt necessary to resolve this problem and a modification of that model is proposed which simulates the data base. Acceleration saturation is achieved by borrowing the idea of the local feedback model for saccades so that one nonlinearity can account for the acceleration-velocity saturation: the main sequence for pursuit. Motor plasticity or motor learning, recently demonstrated for pursuit, is also incorporated and simulated. It was noticed that the offset of pursuit did not show the ringing seen in the onset so this was quantified in one subject. Offset velocity could be characterized by a single exponential with a time constant of about 90 ms. This observation suggests that fixation is not pursuit at zero velocity and that the pursuit system is turned on when needed and off during fixation.

No abstract

Generating, grouping, and labeling free-sketch data is a difficult and time-consuming task for both user study participants and researchers. To simplify this process for both parties, we would like to have users draw isolated shapes instead of complete sketches that must be hand-labeled and grouped, and then use this data to train our free-sketch symbol recognizer. However, it is an open question whether shapes draw in isolation accurately reflect the way users draw shapes in a complete diagram. Furthermore, many of the simplest shape recognition algorithms were designed to recognize gestures, and it is not clear that they will generalize to freely-drawn shapes. To answer these questions, we perform experiments using three different recognizers to measure the effect of the data collection task on recognition accuracy. We find that recognizers trained only on isolated shapes can classify freely-sketched shapes as well as the same recognizers trained on free-sketches. We also show that user-specific training examples significantly improve recognition rates. Finally, we introduce a variant of a popular and simple gesture recognition algorithm that recognizes freely-drawn shapes as well as a highly-accurate but more complex recognizer designed explicitly for free-sketch recognition.