Face Me! Head-Tracker Interface Evaluation on Mobile Devices

Resumen. En este trabajo se presenta el uso de una tarea de selección de objetivos en la evaluación de un head-tracker para dispositivos móviles. Primero, se describe el método de cálculo del rendimiento mediante un ejemplo detallado. A continuación, se discute el método de cálculo para tareas que no cumplen el estándar ISO. Debido a la amplitud no constante de la tarea dentro de cada secuencia, se calcula el rendimiento utilizando dos métodos de agregación de datos: por secuencia, calculando la amplitud media, y por condiciones comunes A-W. Se recomienda calcular el rendimiento utilizando la división de medias y el ajuste de precisión. La media general de rendimiento ha sido de 0,74 bps (un 45 % menor que con una tarea ISO). Se presentan dos recomendaciones de diseño para tareas que no cumplen el estándar ISO: mantener constantes A-W dentro de cada secuencia y utilizar estrategias para evitar el tiempo de reacción.Palabras clave: Ley de Fitts, rendimiento, ISO 9241-411, HCI móvil, head-tracking.Abstract. We used a target-selection task to evaluate head-tracking as an input method for mobile devices. First, the method of calculating Fitts' throughput is described by means of a raw data detailed example. Then, the method of calculating throughput is discussed for non-ISO tasks, since the procedure targets were randomly positioned from trial to trial. Due to a non-constant amplitude within each sequence of trials, throughput was calculated using two methods of data aggregation: the first one, by sequence of trials using the mean amplitude, and the second one, by common A-W conditions.For each data set, we used four methods for calculating throughput. The grand mean for throughput (calculated through the division of means and the adjustment for accuracy) was of 0.74 bps, which is 45 % lower than the value obtained using an ISO task. We recommend to calculate throughput using the division of means plus the adjustment for accuracy, and to avoid using the reciprocal slope of the regression model. We present various design recommendations for non-ISO tasks, such as: i) to keep amplitude and constant target within each sequence of trials, and ii) to use strategies to avoid or remove reaction time.

show abstract

“…In the initial poster [9], the evaluation was limited to selection accuracy, cursor velocity, and selection errors.…”

Section: The "Face Me" Experimentsmentioning

confidence: 99%

Fitts’ Law: On Calculating Throughput and Non-ISO Tasks

Roig-Maimó¹,

MacKenzie²,

Manresa-Yee³

et al. 2018

Rev. colomb. comput.

Self Cite

View full text Add to dashboard Cite

show abstract

“…The Face Me experiment used a mobile head-tracking interface to investigate the e ect of device orientation (portrait, landscape), gain (1.0, 1.5), and target width (88 pixels, 176 px, 212 px). In the initial poster [12], the evaluation was limited to selection accuracy, cursor velocity, and selection errors.…”

Section: The Face Me Experimentsmentioning

confidence: 99%

“…The total number of trials was 19 Participants × 2 Orientations × 2 Gains × 3 Widths × 15 Trials = 3420.A detailed description of the Face Me experiment and the results obtained is found elsewhere[12].…”

mentioning

confidence: 99%

Evaluating fitts' law performance with a non-ISO task

Roig-Maimó

MacKenzie

Manresa-Yee

et al. 2017

Proceedings of the XVIII International Conference on Human Computer Interaction

Self Cite

View full text Add to dashboard Cite

We used a target-selection task to evaluate head-tracking as an input method on a mobile device. The procedure used a non-ISO Fitts' law task since targets were randomly positioned from trial to trial. Due to a non-constant amplitude within each sequence of trials, throughput was calculated using two methods of data aggregation: by sequence of trials using the mean amplitude and by common A-W conditions. For each data set, we used four methods for calculating throughput. The grand mean for throughput calculated using the division of means and the adjustment for accuracy was 0.74 bps, which is 45% lower than the value obtained using an ISO task. We recommend calculating throughput using the division of means (and not the slope reciprocal from the regression model) and with the adjustment for accuracy. We present design recommendation for non-ISO tasks: Keep amplitude and target width constant within each sequence of trials and use strategies to avoid or remove reaction time. CCS CONCEPTS • Human-centered computing → User studies; HCI theory, concepts and models;

show abstract

“…Basically, the head tracking is performed by tracking facial features [ 4 , 27 ] or the entire face (in 2D or 3D) [ 7 , 15 , 28 , 29 ]. Frequently, the facial features selected to track are the eyes [ 27 ] or the nose [ 30 ] and the entire face is usually tracked based on skin color [ 15 , 30 ] or face detectors [ 7 ].…”

Section: Related Workmentioning

confidence: 99%

“…In order to evaluate the effectiveness and efficiency of the camera-based interface, an experiment was conducted to explore its performance with different target sizes (accuracy and velocity), target locations (test if all the screen positions were accessible) and influence of the gain or device orientation [ 30 ]. Due to the small screen size, we could not perform the ISO 9241-9 Multi-directional tapping task [ 36 ].…”

Section: Evaluating the Interface As A Pointing Devicementioning

confidence: 99%

A Robust Camera-Based Interface for Mobile Entertainment

Roig-Maimó

Manresa-Yee

Varona

2016

Sensors

Self Cite

View full text Add to dashboard Cite

Camera-based interfaces in mobile devices are starting to be used in games and apps, but few works have evaluated them in terms of usability or user perception. Due to the changing nature of mobile contexts, this evaluation requires extensive studies to consider the full spectrum of potential users and contexts. However, previous works usually evaluate these interfaces in controlled environments such as laboratory conditions, therefore, the findings cannot be generalized to real users and real contexts. In this work, we present a robust camera-based interface for mobile entertainment. The interface detects and tracks the user’s head by processing the frames provided by the mobile device’s front camera, and its position is then used to interact with the mobile apps. First, we evaluate the interface as a pointing device to study its accuracy, and different factors to configure such as the gain or the device’s orientation, as well as the optimal target size for the interface. Second, we present an in the wild study to evaluate the usage and the user’s perception when playing a game controlled by head motion. Finally, the game is published in an application store to make it available to a large number of potential users and contexts and we register usage data. Results show the feasibility of using this robust camera-based interface for mobile entertainment in different contexts and by different people.

show abstract

Face Me! Head-Tracker Interface Evaluation on Mobile Devices

Cited by 11 publications

References 10 publications

Fitts’ Law: On Calculating Throughput and Non-ISO Tasks

Fitts’ Law: On Calculating Throughput and Non-ISO Tasks

Evaluating fitts' law performance with a non-ISO task

A Robust Camera-Based Interface for Mobile Entertainment

Contact Info

Product

Resources

About