Person-Specific Gaze Estimation from Low-Quality Webcam Images.

Gaze estimation is an established research problem in computer vision. It has various applications in real life, from human-computer interactions to health care and virtual reality, making it more viable for the research community. Due to the significant success of deep learning techniques in other computer vision tasks-for example, image classification, object detection, object segmentation, and object tracking-deep learning-based gaze estimation has also received more attention in recent years. This paper uses a convolutional neural network (CNN) for person-specific gaze estimation. The person-specific gaze estimation utilizes a single model trained for one individual user, contrary to the commonly-used generalized models trained on multiple people's data. We utilized only low-quality images directly collected from a standard desktop webcam, so our method can be applied to any computer system equipped with such a camera without additional hardware requirements. First, we used the web camera to collect a dataset of face and eye images. Then, we tested different combinations of CNN parameters, including the learning and dropout rates. Our findings show that building a person-specific eye-tracking model produces better results with a selection of good hyperparameters when compared to universal models that are trained on multiple users' data. In particular, we achieved the best results for the left eye with 38.20 MAE (Mean Absolute Error) in pixels, the right eye with 36.01 MAE, both eyes combined with 51.18 MAE, and the whole face with 30.09 MAE, which is equivalent to approximately 1.45 degrees for the left eye, 1.37 degrees for the right eye, 1.98 degrees for both eyes combined, and 1.14 degrees for full-face images.

Review and Evaluation of Eye Movement Event Detection Algorithms.

Eye tracking is a technology aimed at understanding the direction of the human gaze. Event detection is a process of detecting and classifying eye movements that are divided into several types. Nowadays, event detection is almost exclusively done by applying a detection algorithm to the raw recorded eye-tracking data. However, due to the lack of a standard procedure for how to perform evaluations, evaluating and comparing various detection algorithms in eye-tracking signals is very challenging. In this paper, we used data from a high-speed eye-tracker SMI HiSpeed 1250 system and compared event detection performance. The evaluation focused on fixations, saccades and post-saccadic oscillation classification. It used sample-by-sample comparisons to compare the algorithms and inter-agreement between algorithms and human coders. The impact of varying threshold values on threshold-based algorithms was examined and the optimum threshold values were determined. This evaluation differed from previous evaluations by using the same dataset to evaluate the event detection algorithms and human coders. We evaluated and compared the different algorithms from threshold-based, machine learning-based and deep learning event detection algorithms. The evaluation results show that all methods perform well for fixation and saccade detection; however, there are substantial differences in classification results. Generally, CNN (Convolutional Neural Network) and RF (Random Forest) algorithms outperform threshold-based methods.

Biometric Identification Based on Keystroke Dynamics.

The purpose of the paper is to study how changes in neural network architecture and its hyperparameters affect the results of biometric identification based on keystroke dynamics. The publicly available dataset of keystrokes was used, and the models with different parameters were trained using this data. Various neural network layers-convolutional, recurrent, and dense-in different configurations were employed together with pooling and dropout layers. The results were compared with the state-of-the-art model using the same dataset. The results varied, with the best-achieved accuracy equal to 82% for the identification (1 of 20) task.

Time Signature Detection: A Survey.

This paper presents a thorough review of methods used in various research articles published in the field of time signature estimation and detection from 2003 to the present. The purpose of this review is to investigate the effectiveness of these methods and how they perform on different types of input signals (audio and MIDI). The results of the research have been divided into two categories: classical and deep learning techniques, and are summarized in order to make suggestions for future study. More than 110 publications from top journals and conferences written in English were reviewed, and each of the research selected was fully examined to demonstrate the feasibility of the approach used, the dataset, and accuracy obtained. Results of the studies analyzed show that, in general, the process of time signature estimation is a difficult one. However, the success of this research area could be an added advantage in a broader area of music genre classification using deep learning techniques. Suggestions for improved estimates and future research projects are also discussed.

Biometric Identification Based on Eye Movement Dynamic Features.

The paper presents studies on biometric identification methods based on the eye movement signal. New signal features were investigated for this purpose. They included its representation in the frequency domain and the largest Lyapunov exponent, which characterizes the dynamics of the eye movement signal seen as a nonlinear time series. These features, along with the velocities and accelerations used in the previously conducted works, were determined for 100-ms eye movement segments. 24 participants took part in the experiment, composed of two sessions. The users' task was to observe a point appearing on the screen in 29 locations. The eye movement recordings for each point were used to create a feature vector in two variants: one vector for one point and one vector including signal for three consecutive locations. Two approaches for defining the training and test sets were applied. In the first one, 75% of randomly selected vectors were used as the training set, under a condition of equal proportions for each participant in both sets and the disjointness of the training and test sets. Among four classifiers: kNN (k = 5), decision tree, naïve Bayes, and random forest, good classification performance was obtained for decision tree and random forest. The efficiency of the last method reached 100%. The outcomes were much worse in the second scenario when the training and testing sets when defined based on recordings from different sessions; the possible reasons are discussed in the paper.

Application of Time-Scale Decomposition of Entropy for Eye Movement Analysis.

The methods for nonlinear time series analysis were used in the presented research to reveal eye movement signal characteristics. Three measures were used: approximate entropy, fuzzy entropy, and the Largest Lyapunov Exponent, for which the multilevel maps (MMs), being their time-scale decomposition, were defined. To check whether the estimated characteristics might be useful in eye movement events detection, these structures were applied in the classification process conducted with the usage of the kNN method. The elements of three MMs were used to define feature vectors for this process. They consisted of differently combined MM segments, belonging either to one or several selected levels, as well as included values either of one or all the analysed measures. Such a classification produced an improvement in the accuracy for saccadic latency and saccade, when compared with the previously conducted studies using eye movement dynamics.

Understanding Eye Movement Signal Characteristics Based on Their Dynamical and Fractal Features.

Eye movement is one of the biological signals whose exploration may reveal substantial information, enabling greater understanding of the biology of the brain and its mechanisms. In this research, eye movement dynamics were studied in terms of chaotic behavior and self-similarity assessment to provide a description of young, healthy, oculomotor system characteristics. The first of the investigated features is present and advantageous for many biological objects or physiological phenomena, and its vanishing or diminishment may indicate a system pathology. Similarly, exposed self-similarity may prove useful for indicating a young and healthy system characterized by adaptability. For this research, 24 young people with normal vision were involved. Their eye movements were registered with the usage of a head-mounted eye tracker, using infrared oculography, embedded in the sensor, measuring the rotations of the left and the right eye. The influence of the preprocessing step in the form of the application of various filtering methods on the assessment of the final dynamics was also explored. The obtained results confirmed the existence of chaotic behavior in some parts of eye movement signal; however, its strength turned out to be dependent on the filter used. They also exposed the long-range correlation representing self-similarity, although the influence of the applied filters on these outcomes was not unveiled.

Implicit Calibration Using Probable Fixation Targets.

Proper calibration of eye movement signal registered by an eye tracker seems to be one of the main challenges in popularizing eye trackers as yet another user-input device. Classic calibration methods taking time and imposing unnatural behavior on eyes must be replaced by intelligent methods that are able to calibrate the signal without conscious cooperation by the user. Such an implicit calibration requires some knowledge about the stimulus a user is looking at and takes into account this information to predict probable gaze targets. This paper describes a possible method to perform implicit calibration: it starts with finding probable fixation targets (PFTs), then it uses these targets to build a mapping-probable gaze path. Various algorithms that may be used for finding PFTs and mappings are presented in the paper and errors are calculated using two datasets registered with two different types of eye trackers. The results show that although for now the implicit calibration provides results worse than the classic one, it may be comparable with it and sufficient for some applications.

An Analysis of Entropy-Based Eye Movement Events Detection.

Analysis of eye movement has attracted a lot of attention recently in terms of exploring areas of people's interest, cognitive ability, and skills. The basis for eye movement usage in these applications is the detection of its main components-namely, fixations and saccades, which facilitate understanding of the spatiotemporal processing of a visual scene. In the presented research, a novel approach for the detection of eye movement events is proposed, based on the concept of approximate entropy. By using the multiresolution time-domain scheme, a structure entitled the Multilevel Entropy Map was developed for this purpose. The dataset was collected during an experiment utilizing the "jumping point" paradigm. Eye positions were registered with a 1000 Hz sampling rate. For event detection, the knn classifier was applied. The best classification efficiency in recognizing the saccadic period ranged from 83% to 94%, depending on the sample size used. These promising outcomes suggest that the proposed solution may be used as a potential method for describing eye movement dynamics.

Searching for Chaos Evidence in Eye Movement Signals.

Most naturally-occurring physical phenomena are examples of nonlinear dynamic systems, the functioning of which attracts many researchers seeking to unveil their nature. The research presented in this paper is aimed at exploring eye movement dynamic features in terms of the existence of chaotic nature. Nonlinear time series analysis methods were used for this purpose. Two time series features were studied: fractal dimension and entropy, by utilising the embedding theory. The methods were applied to the data collected during the experiment with "jumping point" stimulus. Eye movements were registered by means of the Jazz-novo eye tracker. One thousand three hundred and ninety two (1392) time series were defined, based on the horizontal velocity of eye movements registered during imposed, prolonged fixations. In order to conduct detailed analysis of the signal and identify differences contributing to the observed patterns of behaviour in time scale, fractal dimension and entropy were evaluated in various time series intervals. The influence of the noise contained in the data and the impact of the utilized filter on the obtained results were also studied. The low pass filter was used for the purpose of noise reduction with a 50 Hz cut-off frequency, estimated by means of the Fourier transform and all concerned methods were applied to time series before and after noise reduction. These studies provided some premises, which allow perceiving eye movements as observed chaotic data: characteristic of a space-time separation plot, low and non-integer time series dimension, and the time series entropy characteristic for chaotic systems.

Simplified automatic method for measuring the visual field using the perimeter ZERK 1.

BACKGROUND: Currently available perimeters have limited capabilities of performing measurements of the visual field in children. In addition, they do not allow for fully automatic measurement even in adults. The patient in each case (in any type of perimeter) has at his disposal a button which he uses to indicate that he has seen a light stimulus. Such restrictions have been offset in the presented new perimeter ZERK 1. METHODS: The paper describes a new type of automated, computerized perimeter designed to test the visual field in children and adults. The new perimeter and proprietary software enable to carry out tests automatically (without the need to press any button). The presented full version of the perimeter has been tested on a head phantom. The next steps will involve clinical trials and a comparison with measurements obtained using other types of perimeters. RESULTS: The perimeter ZERK 1 enables automatic measurement of the visual field in two axes (with a span of 870 mm and a depth of 525 mm) with an accuracy of not less than 1(o) (95 LEDs on each arm) at a typical position of the patient's head. The measurement can be carried out in two modes: default/typical (lasting about 1 min), and accurate (lasting about 10 min). Compared with available and known types of perimeters, it has an open canopy, proprietary software and cameras tracking the eye movement, automatic control of fixation points, light stimuli with automatically preset light stimulus intensity in the following ranges: 550-700 mcd (red 620-630 nm), 1100-1400 mcd (green 515-530 nm), 200-400 mcd (blue 465-475 nm). CONCLUSIONS: The paper presents a new approach to the construction of perimeters based on automatic tracking of the eye movements in response to stimuli. The unique construction of the perimeter and the software allow for its mobile use in the examination of children and bedridden patients.

METER2800: A novel dataset for music time signature detection.

The Meter2800 dataset is an important contribution to Music Information Retrieval (MIR) research, as it is the first dataset to include audio files specifically designed for time signature detection. By combining audio files from three renowned datasets and including additional tracks, we have created a comprehensive and diverse collection of 2800 audio tracks that overcomes the limitations of existing audio datasets. The dataset includes 2.26GB of high-quality audio, which has been annotated with metadata, pre-computed features, tempo and time signature. In addition, we propose a train/test split and provide baseline results for time signature detection. The dataset is freely available for the research community and is available online for download. We believe that Meter2800 will contribute to the advancement of Music Information Retrieval research, particularly in the area of time signature detection. In technical validation, four classification experiments were conducted using four types of machine learning algorithms: SVM, KNN, Naive Bayes, and Random Forest.

Datasets for learning of unknown characteristics of dynamical systems.

The ability to uncover characteristics based on empirical measurement is an important step in understanding the underlying system that gives rise to an observed time series. This is especially important for biological signals whose characteristic contributes to the underlying dynamics of the physiological processes. Therefore, by studying such signals, the physiological systems that generate them can be better understood. The datasets presented consist of 33,000 time series of 15 dynamical systems (five chaotic and ten non-chaotic) of the first, second, or third order. Here, the order of a dynamical system means its dimension. The non-chaotic systems were divided into the following classes: periodic, quasi-periodic, and non-periodic. The aim is to propose datasets for machine learning methods, in particular deep learning techniques, to analyze unknown dynamical system characteristics based on obtained time series. In technical validation, three classifications experiments were conducted using two types of neural networks with long short-term memory modules and convolutional layers.

Application of eye tracking in medicine: A survey, research issues and challenges.

The performance and quality of medical procedures and treatments are inextricably linked to technological development. The application of more advanced techniques provides the opportunity to gain wider knowledge and deeper understanding of the human body and mind functioning. The eye tracking methods used to register eye movement to find the direction and targets of a person's gaze are well in line with the nature of the topic. By providing methods for capturing and processing images of the eye it has become possible not only to reveal abnormalities in eye functioning but also to conduct cognitive studies focused on learning about peoples' emotions and intentions. The usefulness of the application of eye tracking technology in medicine was proved in many research studies. The aim of this paper is to give an insight into those studies and the way they utilize eye imaging in medical applications. These studies were differentiated taking their purpose and experimental paradigms into account. Additionally, methods for eye movement visualization and metrics for its quantifying were presented. Apart from presenting the state of the art, the aim of the paper was also to point out possible applications of eye tracking in medicine that have not been exhaustively investigated yet, and are going to be a perspective long-term direction of research.

Vision Diagnostics and Treatment System for Children with Disabilities.

Vision plays a crucial role in children's mental development. Therefore, early diagnosis of any vision disparities and implementation of a correct therapy is very important. However, carrying out such a procedure in case of young children and especially children with brain dysfunctions poses some limitations due to cooperation problems. The vision diagnostics and treatment (VisDaT) system presented in this paper is meant to help therapists in proper diagnosis and treatment involving such children. It utilizes a computer connected to two monitors and equipped with a specialized software. The main system components are as follows: an eye tracker recording child's eye movements and a digital camera monitoring online child's reactions. The system is equipped with a specialized software, which creates the opportunity to stimulate children's vision with a dedicated stimulus and post hoc analyses of recorded sessions, which enable making decision as to the future treatment.

Gaze Self-Similarity Plot - A New Visualization Technique.

Eye tracking has become a valuable way for extending knowledge of human behavior based on visual patterns. One of the most important elements of such an analysis is the presentation of obtained results, which proves to be a challenging task. Traditional visualization techniques such as scan-paths or heat maps may reveal interesting information, nonetheless many useful features are still not visible, especially when temporal characteristics of eye movement is taken into account. This paper introduces a technique called gaze self-similarity plot (GSSP) that may be applied to visualize both spatial and temporal eye movement features on the single two-dimensional plot. The technique is an extension of the idea of recurrence plots, commonly used in time series analysis. The paper presents the basic concepts of the proposed approach (two types of GSSP) complemented with some examples of what kind of information may be disclosed and finally showing areas of the GSSP possible applications.