Task classification model for visual fixation, exploration, and search

@article{Kumar2019TaskCM,
  title={Task classification model for visual fixation, exploration, and search},
  author={Ayush Kumar and Anjul Tyagi and Michael Burch and Daniel Weiskopf and Klaus Mueller},
  journal={Proceedings of the 11th ACM Symposium on Eye Tracking Research \& Applications},
  year={2019}
}
  • Ayush Kumar, Anjul Tyagi, K. Mueller
  • Published 25 June 2019
  • Computer Science
  • Proceedings of the 11th ACM Symposium on Eye Tracking Research & Applications
Yarbus' claim to decode the observer's task from eye movements has received mixed reactions. In this paper, we have supported the hypothesis that it is possible to decode the task. We conducted an exploratory analysis on the dataset by projecting features and data points into a scatter plot to visualize the nuance properties for each task. Following this analysis, we eliminated highly correlated features before training an SVM and Ada Boosting classifier to predict the tasks from this filtered… 

Figures and Tables from this paper

Task Classification During Visual Search Using Classic Machine Learning and Deep Learning

TLDR
This research runs machine learning and deep learning algorithms to identify the task type from eye-tracking data and takes a “visual” approach by experimenting on variations of Computer Vision algorithms like Convolutional Neural Networks on the visual representations of the user gaze scan paths.

Challenges in Interpretability of Neural Networks for Eye Movement Data

TLDR
The main problems for interpreting neural networks with this type of data are characterized, how they differ from the problems faced in other domains, and why existing techniques are not sufficient to address all of these issues.

GazeMAE: General Representations of Eye Movements using a Micro-Macro Autoencoder

TLDR
This work proposes an abstract representation of eye movements that preserve the important nuances in gaze behavior while being stimuli-agnostic, and accurately discriminates between gender and age groups and outperforms previous works on biometrics and stimuli classification.

Eye Gaze Estimation Model Analysis

TLDR
The purpose of this work is to discuss various model types for eye gaze estimation and present the results from predicting gaze direction using eye landmarks in unconstrained settings.

NAS-Navigator: Visual Steering for Explainable One-Shot Deep Neural Network Synthesis.

TLDR
NAS-Navigator is presented, a visual analytics (VA) system aiming to solve three problems with One-Shot NAS; explainability, HIL design, and performance improvements compared to existing state-of-the-art (SOTA) techniques.

An Open Source Interactive Visual Analytics Tool for Comparative Programming Comprehension

This paper proposes a visual analytics tool consisting of several views and perspectives on eye movement data collected during code tasks when focus of this work is on code and program comprehension.

Visual Steering for One-Shot Deep Neural Network Synthesis

TLDR
This paper presents a framework that allows analysts to effectively build the solution sub-graph space and guide the network search by injecting their domain knowledge and applies this technique in an iterative manner to converge to the best performing neural network architecture for a given application.

PC-Expo: A Metrics-Based Interactive Axes Reordering Method for Parallel Coordinate Displays

TLDR
PC-Expo is presented, a real-time visual analytics framework for all-in-one PCP line pattern detection and axes reordering, and expands prior work on PCP axes re ordering by developing real- time, local detection schemes for the 12 most common analysis tasks (properties).

ICE: An Interactive Configuration Explorer for High Dimensional Categorical Parameter Spaces

TLDR
This work proposes a novel approach, the Interactive Configuration Explorer (ICE), which directly addresses the need of analysts to learn how the dependent numerical variable is affected by the parameter settings given multiple optimization objectives.

Infographics Wizard: Flexible Infographics Authoring and Design Exploration

TLDR
Evaluation results confirm that by using the semi‐automated infographic framework, designers from all expertise levels can generate generic infographic designs faster than existing methods while maintaining the same quality as hand‐designed infographics templates.

References

SHOWING 1-10 OF 18 REFERENCES

Reconsidering Yarbus: A failure to predict observers’ task from eye movement patterns

Defending Yarbus: eye movements reveal observers' task.

TLDR
Yarbus's idea that human eye-movement patterns are modulated top down by different task demands is supported by the data and continues to be an inspiration for future computational and experimental eye- Movement research.

Saccades and microsaccades during visual fixation, exploration, and search: foundations for a common saccadic generator.

TLDR
The results show that microsaccades are produced during the fixation periods that occur during visual exploration and visual search, and support the hypothesis of a common oculomotor generator for saccades and micros Accades.

Visual Multi-Metric Grouping of Eye-Tracking Data

TLDR
An algorithmic and visual grouping of participants and eye-tracking metrics derived from recorded eye- tracking data using a similarity matrix visualization to visually represent the affine combination of metrics utilizing an algorithmic grouping of subjects that leads to distinct visual groups of similar behavior is presented.

Support-Vector Networks

TLDR
High generalization ability of support-vector networks utilizing polynomial input transformations is demonstrated and the performance of the support- vector network is compared to various classical learning algorithms that all took part in a benchmark study of Optical Character Recognition.

Popular Ensemble Methods: An Empirical Study

TLDR
This work suggests that most of the gain in an ensemble's performance comes in the first few classifiers combined; however, relatively large gains can be seen up to 25 classifiers when Boosting decision trees.

Experiments with a New Boosting Algorithm

TLDR
This paper describes experiments carried out to assess how well AdaBoost with and without pseudo-loss, performs on real learning problems and compared boosting to Breiman's "bagging" method when used to aggregate various classifiers.

Comparison of classifier methods: a case study in handwritten digit recognition

  • L. BottouCorinna Cortes V. Vapnik
  • Computer Science
    Proceedings of the 12th IAPR International Conference on Pattern Recognition, Vol. 3 - Conference C: Signal Processing (Cat. No.94CH3440-5)
  • 1994
This paper compares the performance of several classifier algorithms on a standard database of handwritten digits. We consider not only raw accuracy, but also training time, recognition time, and

A comparison of methods for multiclass support vector machines

TLDR
Decomposition implementations for two "all-together" multiclass SVM methods are given and it is shown that for large problems methods by considering all data at once in general need fewer support vectors.

Multidimensional scaling by optimizing goodness of fit to a nonmetric hypothesis

TLDR
The fundamental hypothesis is that dissimilarities and distances are monotonically related, and a quantitative, intuitively satisfying measure of goodness of fit is defined to this hypothesis.