Modeling Touch Point Distribution with Rotational Dual Gaussian Model.

Area of Science:

• Human-Computer Interaction
• Usability Engineering
• Touchscreen Technology

Background:

• Touch point distribution models are crucial for designing effective touchscreen interfaces.
• Existing models like the Dual Gaussian model do not fully account for finger movement direction.
• Understanding touch point distribution is key to optimizing user interaction and performance.

Purpose of the Study:

• To investigate the effect of finger movement direction on touch point distribution.
• To propose a refined model, the Rotational Dual Gaussian model, that incorporates finger movement direction.
• To evaluate the performance of the proposed model against existing methods.

Main Methods:

• Developed the Rotational Dual Gaussian model, where the major axis aligns with finger movement direction.
• Introduced the use of projected target dimensions instead of nominal ones.
• Evaluated the model on three empirical datasets for touch point distribution prediction.

Main Results:

• The Rotational Dual Gaussian model accurately reflects the elongation of touch point distribution along the finger movement direction.
• The new model significantly outperforms the original Dual Gaussian Model in prediction tests.
• Reduced the Root Mean Square Error (RMSE) of touch error rate prediction from 8.49% to 4.95%.

Conclusions:

• The Rotational Dual Gaussian model provides a more accurate prediction of touch point distribution, especially considering finger movement.
• This model enhances target acquisition accuracy and improves soft keyboard decoding on devices like smartwatches.
• Incorporating directional information in touch modeling leads to significant performance gains.