Search

Autism Spectrum Disorder (ASD) is a neuro-developmental disorder that limits social and cognitive abilities. ASD has no cure so early diagnosis is important for reducing its impact. The current behavioral observation-based subjective-diagnosis systems (e.g., DSM-5 or ICD-10) frequently misdiagnose subjects. Therefore, researchers are attempting to develop automated diagnosis systems with minimal human intervention, quicker screening time, and better outreach. This paper is a PRISMA-based systematic review examining the potential of automated autism detection system with Human Activity Analysis (HAA) to look for distinctive ASD characteristics such as repetitive behavior, abnormal gait and visual saliency. The literature from 2011 onward is qualitatively and quantitatively analyzed to investigate whether HAA can identify the features of ASD, the level of its classification accuracy, the degree of human intervention, and screening time. Based on these findings, we discuss the approaches, challenges, resources, and future directions in this area. According to our quantitative assessment of the dataset Zunino et al. (IEEE: 3421–3426, 2018 [1]), Inception v3 and LSTM Zunino et al. (IEEE: 3421–3426, 2018 [1]) give the highest accuracy (89%) for repetitive behavior. For abnormal gait-based approach, the multilayer perceptron gives 98% accuracy based on 18 features from dataset Abdulrahman et al. (COMPUSOFT: An International Journal of Advanced Computer Technology 9(8):3791–3797, 2020 [2]). For gaze pattern, a saliency-metric feature-based learning Rahman et al. (Int Conf Pattern Recognit, 2020 [3]) gives 99% accuracy on dataset Duan et al. (Proceedings of the 10th ACM Multimedia Systems Conference: 255–260, 2019 [4]), while an algorithm involving statistical features and Decision Trees yields an accuracy of 76% on dataset Yaneva et al. (Proceedings of the Internet of Accessible Things. W4A '18, Association for Computing Machinery, New York, NY, USA, 1–10, 2018 [5]). In terms of the state of the art, fully automated HAA systems for ASD diagnosis show promise but are still in developmental stages. However, this is an active research field, and HAA has good prospects for helping to diagnose ASD objectively in less time with better accuracy. [ABSTRACT FROM AUTHOR]

Image feature point algorithms and their associated regional descriptors can be viewed as primitive detectors of visually salient information. In this paper, a new method for constructing a visual attention probability map using features is proposed. (Throughout this work, we use SURF features yet the algorithm is not limited to SURF alone.). This technique is validated using comprehensive human eye-tracking experiments. We call this algorithm 'visual interest' (VI) since the resultant segmentation reveals image regions that are visually salient during the performance of multiple observer search tasks. We demonstrate that it works on generic, eye-level photographs and is not dependent on heuristic tuning. We further show that the descriptor-matching property of the SURF feature points can be exploited via object recognition to modulate the context of the attention probability map for a given object search task, refining the salient area. We fully validate the VI algorithm through applying it to salient compression using a pre-blur of non-salient regions prior to JPEG and conducting comprehensive observer performance tests. When using the object contextualisation, we conclude that JPEG files are around 33 % larger than they need to be to fully represent the task-relevant information within them. We finally demonstrate the utility of the segmentation as a region of interest in JPEG2000 compression to achieve superior image quality (measured statistically using PSNR and SSIM) over the automatically selected salient image regions while reducing the image filesize by down to 25 % of that of the original. Our technique therefore delivers superior compression performance through the detection and selective preservation of visually salient information relevant to multiple observer tasks. In contrast to the state of the art in task-directed visual attention models, the VI algorithm reacts only to the image content and requires no detailed prior knowledge of the scene nor of the ultimate observer task. [ABSTRACT FROM AUTHOR]

In this article, a conceptual framework developed to acquire expert knowledge from eye-tracking data of skilled individuals is presented. Domain-specific knowledge is acquired from the visual behaviour of subjects whose eye movements are recorded while solving complex visual tasks. It is argued that relevant insights into the cognitive strategies followed by the observers to solve the visual search tasks may be gained by analysing the eye-tracking data in generic feature spaces, which are at the basis of the selected scheme for knowledge representation. In this context, a feature space is a domain in which each dimension is defined as a mathematical construct, which may correspond to perceptually meaningful visual cues and which can take either numerical or categorical values. A special case of such feature spaces is the spatial domain in which the spatial coordinates of the gaze points define the dimensions of such domain. In the proposed conceptual framework, the definition of similarities between visual search patterns is essential to characterise the stereotypical visual behaviour of a group of observers, and thus expert knowledge. Furthermore, since knowledge representation is closely related to the feature domain in which the search is analysed, feature relevance measures become central to knowledge gathering, and the main aspects regarding their definition are discussed in this work. Following a detailed presentation of the conceptual framework, a practical application dealing with expert knowledge gathering in lung radiology is shown both as a proof of concept and also to illustrate a particular functional implementation of the framework. [ABSTRACT FROM AUTHOR]

video sequence is more than a sequence of still images. It contains a strong spatial-temporal correlation between the regions of consecutive frames. The most important characteristic of videos is the perceived motion foreground objects across the frames. The motion of foreground objects dramatically changes the importance of the objects in a scene and leads to a different saliency map of the frame representing the scene. This makes the saliency analysis of videos much more complicated than that of still images. In this paper, we investigate saliency in video sequences and propose a novel spatiotemporal saliency model devoted for video surveillance applications. Compared to classical saliency models based on still images, such as Itti's model, and space-time saliency models, the proposed model is more correlated to visual saliency perception of surveillance videos. Both bottom-up and top-down attention mechanisms are involved in this model. Stationary saliency and motion saliency are, respectively, analyzed. First, a new method for background subtraction and foreground extraction is developed based on content analysis of the scene in the domain of video surveillance. Then, a stationary saliency model is setup based on multiple features computed from the foreground. Every feature is analyzed with a multi-scale Gaussian pyramid, and all the features conspicuity maps are combined using different weights. The stationary model integrates faces as a supplement feature to other low level features such as color, intensity and orientation. Second, a motion saliency map is calculated using the statistics of the motion vectors field. Third, both motion saliency map and stationary saliency map are merged based on center-surround framework defined by an approximated Gaussian function. The video saliency maps computed from our model have been compared to the gaze maps obtained from subjective experiments with SMI eye tracker for surveillance video sequences. The results show strong correlation between the output of the proposed spatiotemporal saliency model and the experimental gaze maps. [ABSTRACT FROM AUTHOR]

The problem of predicting where people look at, or equivalently salient region detection, has been related to the statistics of several types of low-level image features. Among these features, contrast and edge information seem to have the highest correlation with the fixation locations. The contrast distribution of natural images can be adequately characterized using a two-parameter Weibull distribution. This distribution catches the structure of local contrast and edge frequency in a highly meaningful way. We exploit these observations and investigate whether the parameters of the Weibull distribution constitute a simple model for predicting where people fixate when viewing natural images. Using a set of images with associated eye movements, we assess the joint distribution of the Weibull parameters at fixated and non-fixated regions. Then, we build a simple classifier based on the log-likelihood ratio between these two joint distributions. Our results show that as few as two values per image region are already enough to achieve a performance comparable with the state-of-the-art in bottom-up saliency prediction. [ABSTRACT FROM AUTHOR]