Off-line Accuracy Research Articles

The effect of sentence length on question comprehension in children with cochlear implants

Objectives The present study investigated the comprehension of subject and object who and which questions in children with cochlear implants (CI). Methods Growth Curve Analysis (GCA) was used to compare eye gaze fixations and gaze patterns to the appropriate subject or object nouns within a four-picture array between 16 children with CI and 31 children with typical hearing (aged 7;0-12;0) on wh-questions with and without added adjectives to increase length. Offline accuracy was also compared. Results Findings indicated children with typical hearing exhibited more fixations to the target noun across all conditions, supporting higher comprehension accuracy. Both groups of children demonstrated more fixations to the target noun in object questions and questions without added length. Patterns of eye movement were significantly different between groups, suggesting different patterns of eye gaze across the array before fixation on the target noun. Children with CI exhibited fewer fixations, slower speed to fixation, and differences in gaze patterns that may imply the presence of processing limitations. Error analyses also suggested that children with CI frequently fixated on a picture similar to the target noun. Conclusions Results indicate children with CI comprehend questions more slowly than their hearing peers, which may be related to limitations in working memory.

Flexible Optical Fiber-Based Smart Textile Sensor for Human–Machine Interaction

Flexible sensors are the main components of wearable devices. Compared with flexible substrate-based skin-like sensors, flexible textile sensors have the intrinsic advantages of excellent breathability, compliance with the human body and comfort to human skin. In this article, a flexible optical fiber-based smart textile (FOFT) sensor has been proposed for the application of intelligent human–machine interaction (HMI) and healthcare. The FOFT sensor integrated fiber Bragg gratings (FBGs) with yarns in the way of the plain weave. A double-layer flexible optical fiber crossover structure is proposed to improve the measurement accuracy of the sensor for human body signals. Such a structure leads to a highly flexible textile and wearing comfort. The weak characteristic information of the human body can be extracted by the sensor based on the wavelength demodulation of double FBGs. Combined with the backpropagation (BP) neural network, an FOFT sensor-based gesture recognition application with an off-line accuracy of 97.02% has been developed. Furthermore, the FOFT sensor is used to monitor the human respiratory signals under different postures and obtain the respiratory rate. A series of experiments demonstrate that the FOFT sensor has a huge potential in HMI and healthcare.

Improve the robustness of data mining algorithm against adversarial evasion attack

Conventional data mining theories developed for general-purpose applications commonly focus on the reducing the bias and variance on the ideal i.i.d. datasets, but neglecting its potential failure on maliciously generated data points by observing the system's behaviours. Therefore, dealing with these adversarial samples is an essential part of a security system to handle the data that are intentionally made to deceive the system. Due to this concern, this paper proposes a novel approach that introduces uncertainty to the model behaviour, in order to obfuscate the decision process of the attacking strategy and improve the robustness of security system against attacks that try to evade the detection. Our approach addresses three problems. First, we build a pool of mining models to improve robustness of a variety of mining algorithms, similar to ensemble learning but focusing on the optimisation the trade-off between off-line accuracy and robustness. Second, we randomly select a subset of models at run time (when the model is used for detection) to further boost the robustness. Third, we propose a theoretical framework that bounds the minimal number of features an attacker needs to modify given a set of selected models.

The processing of English which-questions in adult L2 learners: Effects of L1 transfer and proficiency

Abstract In a visual-world eye-tracking study, we investigate the on-line comprehension of subject and object which-questions in 60 German-English adult second-language (L2) learners at various stages of proficiency. In particular, we examine whether adult L2 learners follow the same acquisitional trajectories as children or whether L1 transfer dictates a different development. Previous research on monolingual English children shows that child learners initially commit to a strong subject preference and use inflectional cues, i.e. number marking, to overcome structural biases (e.g. Contemori and Marinis 2014b). The adult data on comprehension accuracy show strong effects of proficiency. Intermediate-proficiency L2 learners have an overwhelming structural preference for subject questions, and they are not sensitive to inflectional cues that disambiguate towards object readings. High-intermediate learners come to exploit inflectional cues for revising an initial subject-preference to the target object reading. Finally, advanced learners display native-like comprehension patterns. In contrast to the off-line accuracy data, the analysis of the eye movements shows that all groups of L2 adults make incremental use of inflectional cues in on-line comprehension. These findings indicate that inflectional cues are integrated in the course of L2 development in similar ways as in monolingual development. We discuss the results in the context of current approaches to L2 processing and child-adult differences in language development.

Offline accuracy: A potentially misleading metric in myoelectric pattern recognition for prosthetic control.

Offline accuracy has been the preferred performance measure in myoelectric pattern recognition (MPR) for the prediction of motion volition. In this study, different metrics relating the fundamental binary prediction outcomes were analyzed. Our results indicate that global accuracy is biased by 1) the unbalanced number of possible true positive and negative outcomes, and 2) the almost perfect specificity and negative predicted value, which were consistently found across algorithms, topologies, and movements (individual and simultaneous). Therefore, class-specific accuracy is advisable instead. Additionally, we propose the use of precision (positive predictive value) and sensitivity (recall) as a complement to accuracy to better describe the discrimination capabilities of MPR algorithms, as these consider the effect of false predictions. However, all the studied offline metrics failed to predict real-time decoding, and therefore real-time testing continue to be necessary to truly evaluate the clinical usability of MPR.

Realtime recognition of multi-finger prehensile gestures

This paper presents an on-line myoelectric control system which can classify eight prehensile hand gestures with only two electrodes. The overlapping windowing scheme is adopted in the system leading a continuous decisions flow. We choose mean absolute value (MAV), variance (VAR), the fourth-order autoregressive (AR) coefficient and Sample entropy (SampEn) as the feature set and utilize the linear discriminant analysis (LDA) to reduce the dimension and obtain the projected feature sets. The current projected feature set and the previous one are “pre-smoothed” before the classification, and then a decision is generated by LDA classifier. To get the final decision from the decisions flow, the current decision and m previous decisions are “post-smoothed”. The method mentioned above can obtain a 99.04% off-line accuracy rate and a 97.35% on-line accuracy rate for individual gesture. By choosing a proper value of m, this method can also get a 99.79% accuracy rate for on-line recognition of complex sequences of hand gestures without interruption. In addition, a virtual hand has been developed to display the on-line recognition result visually, and a proper control strategy is proposed to realize the continuous switch of hand gestures.

The impact of number mismatch and passives on the real-time processing of relative clauses

Language processing plays a crucial role in language development, providing the ability to assign structural representations to input strings (e.g., Fodor, 1998). In this paper we aim at contributing to the study of children's processing routines, examining the operations underlying the auditory processing of relative clauses in children compared to adults. English-speaking children (6;0-8;11) and adults participated in the study, which employed a self-paced listening task with a final comprehension question. The aim was to determine (i) the role of number agreement in object relative clauses in which the subject and object NPs differ in terms of number properties, and (ii) the role of verb morphology (active vs. passive) in subject relative clauses. Even though children's off-line accuracy was not always comparable to that of adults, analyses of reaction times results support the view that children have the same structural processing reflexes observed in adults.

A Large Clinical Study on the Ability of Stroke Patients to Use an EEG-Based Motor Imagery Brain-Computer Interface

Brain-computer interface (BCI) technology has the prospects of helping stroke survivors by enabling the interaction with their environ ment through brain signals rather than through muscles, and restoring motor function by inducing activity-dependent brain plasticity. This paper presents a clinical study on the extent of detectable brain signals from a large population of stroke patients in using EEG-based motor imagery BCI. EEG data were collected from 54 stroke patients whereby finger tapping and motor imagery of the stroke-affected hand were performed by 8 and 46 patients, respectively. EEG data from 11 patients who gave further consent to perform motor imagery were also collected for second calibration and third independent test sessions conducted on separate days. Off-line accuracies of classifying the two classes of EEG from finger tapping or motor imagery of the stroke-affected hand versus the EEG from background rest were then assessed and compared to 16 healthy subjects. The mean off-line accuracy of detecting motor imagery by the 46 patients (mu=0.74) was significantly lower than finger tapping by 8 patients (mu=0.87, p=0.008), but not significantly lower than motor imagery by healthy subjects (mu=0.78, p=0.23). Six stroke patients performed motor imagery at chance level, and no correlation was found between the accuracies of detecting motor imagery and their motor impairment in terms of Fugl-Meyer Assessment (p=0.29). The off-line accuracies of the 11 patients in the second session (mu=0.76) were not significantly different from the first session (mu=0.72, p=0.16), or from the on-line accuracies of the third independent test session (mu=0.82, p=0.14). Hence this study showed that the majority of stroke patients could use EEG-based motor imagery BCI.

Developmental Differences in Visual and Auditory Processing of Complex Sentences

Children aged 8 through 11 (N = 250) were given a word-by-word sentence task in both the visual and auditory modes. The sentences included an object relative clause, a subject relative clause, or a conjoined verb phrase. Each sentence was followed by a true-false question, testing the subject of either the first or second verb. Participants were also given two memory span measures: digit span and reading span. High digit span children slowed down more at the transition from the main to the relative clause than did the low digit span children. The findings suggest the presence of a U-shaped learning pattern for on-line processing of restrictive relative clauses. Off-line accuracy scores showed different patterns for good comprehenders and poor comprehenders. Poor comprehenders answered the second verb questions at levels that were consistently below chance. Their answers were based on an incorrect local attachment strategy that treated the second noun as the subject of the second verb. For example, they often answered yes to the question "The girl chases the policeman" after the object relative sentence "The boy that the girl sees chases the policeman." Interestingly, low memory span poor comprehenders used the local attachment strategy less consistently than high memory span poor comprehenders, and all poor comprehenders used this strategy less consistently for harder than for easier sentences.