Rotation,Translation and Scale Invariant Sign Word Recognition Using Deep Learning

Communication between people with disabilities and people who do not understand sign language is a growing social need and can be a tedious task.One of the main functions of sign language is to communicate with each other through hand gestures.Recognition of hand gestures has become an important challenge for the recognition of sign language.There are many existing models that can produce a good accuracy,but if the model test with rotated or translated images,they may face some difficulties to make good performance accuracy.To resolve these challenges of hand gesture recognition,we proposed a Rotation,Translation and Scale-invariant sign word recognition system using a convolu-tional neural network(CNN).We have followed three steps in our work:rotated,translated and scaled(RTS)version dataset generation,gesture segmentation,and sign word classification.Firstly,we have enlarged a benchmark dataset of 20 sign words by making different amounts of Rotation,Translation and Scale of the ori-ginal images to create the RTS version dataset.Then we have applied the gesture segmentation technique.The segmentation consists of three levels,i)Otsu Thresholding with YCbCr,ii)Morphological analysis:dilation through opening morphology and iii)Watershed algorithm.Finally,our designed CNN model has been trained to classify the hand gesture as well as the sign word.Our model has been evaluated using the twenty sign word dataset,five sign word dataset and the RTS version of these datasets.We achieved 99.30%accuracy from the twenty sign word dataset evaluation,99.10%accuracy from the RTS version of the twenty sign word evolution,100%accuracy from thefive sign word dataset evaluation,and 98.00%accuracy from the RTS versionfive sign word dataset evolution.Furthermore,the influence of our model exists in competitive results with state-of-the-art methods in sign word recognition.

Hybrid Features for an Arabic Word Recognition System

This research proposes and implements an Arabic Sub-Words Recognition System （ASWR）. The system focuses on employing a combination of statistical and structural features to provide complete pattern＇s description and enhances the recognition rate. Support Vector Machines （SVMs） is utilized as a promising pattern recognition tool. In addition to that, the problems of dots and holes are solved in a completely different way from the ones previously employed. The proposed system proceeds in several phases as follows： （1） image acquisition, （2） binarisation, （3） morphological processing, （4） feature extraction, which includes statistical features, i.e., moment invariants, and structural features, i.e., dot number, dot position, and number of holes, features, and （5） classification, using multi-class SVMs and applying a one-against-all technique. The proposed system has been tested using different sets of words and subwords and has achieved a nearly 98.90% recogiaition rate. Comparative results with NNs are also presented.

How Word Recognition Is Affected by Schema Activation:An Investigation Through Listening

单词辨识是听力过程中的基本环节。本文探讨了听前背景图式知识的激活会对单词辨识产生的影响。研究发现,与听力前没有激活相关图式知识的控制组相比,图式激活的实验组同学不仅可以在听力中辨识更多的单词,而且他们可以识别更多在语流中读音发生变化的单词;更好避免因某个音素相似而造成的单词误听;减少因已理解的内容无法接纳某词而造成拒绝接受该词的可能性;同时也降低了因对已听内容的回顾而造成无法辨识后续单词的几率。

A Case Study on Intermediate CSL Learners' Word Recognition Processes and Strategies in Contextual Reading Settings

This study investigates word recognition processes and strategies of intermediate learners of Chinese as a Second Language （CSL） in contextual reading settings. Two intermediate CSL learners were chosen as research participants, and think-aloud methods and retrospective interviews were used to collect data. The data were analyzed by using Moustakas＇ data analysis procedure, CresweU＇s three steps and Bogdon and Biklen＇s data analysis methods. Results indicated that intermediate CSL learners go through different processes of word recognition as it might be automatic, based on context, pronunciation, previous knowledge and the meaning of characters, or, in case of word recognition failure, skipping the words or skipping them but reading them again later; and their word recognition strategies in contextual reading settings mainly include cognitive strategies and self-regulatory strategies. Among these strategies, cognitive strategies consist of direct transformation, translation, interpretation, guessing, inferring and finding key words; and self-regulatory strategies include metacognitive strategies, behavior regulating strategies, emotion regulating strategies and motivation regulating strategies. A model of intermediate CSL learners＇ word recognition strategies can be constructed based on the results. The present study provides both theoretical and pedagogical implications in the field of CSL vocabulary acquisition and teaching.

Characteristics of Mandarin Open-set Word Recognition Development among Chinese Children with Cochlear Implants

Background： Cochlear implants （Cls） can improve speech recognition for children with severe congenital hearing loss, and open-set word recognition is an important efficacy measure. This study examined Mandarin open-set word recognition development among Chinese children with Cls and normal hearing （NH）. Methods： This study included 457 children with CIs and 131 children with NH, who completed the Mandarin lexical neighborhood test. The results for children at 1-8 years alter receiving their Cls were compared to those from the children with NH using linear regression analysis and analysis of variance. Results： Recognition of disyllabic easy words, disyllabic hard words, monosyllabic easy words, and monosyllabic hard words increased with time after CI implantation. Scores for cases with implantation before 3 years old were significantly better than those for implantation after 3 years old. There were significant differences in open-set word recognition between the CI and NH groups. For implantation before 2 years, there was no significant difference in recognition at the ages of 6-7 years, compared to 3-year-old children with NH, or at the age of 10 years, compared to 6-year-old children with NH. For implantation before 3 years, there was no significant difference in recognition at the ages of 8 9 years, compared to 3-year-old children with NH, or at the age of 10 years, compared to 6-year-old children with NH. For implantation after 3 years, there was a significant difference in recognition at the age of 13 years, compared to 3-year-old children with NH. Conclusions： Mandarin open-set word recognition increased with time after CI implantation, and the age at implantation had a significant effect on long-term speech recognition. Chinese children with Cls had delayed but similar development of recognition, compared to norrnal children. Early CI implantation can shorten the gap between children with Cls and normal children.

Improvement in word recognition following treatment failure for sudden sensorineural hearing loss

Objectives:Patients with sudden sensorineural hearing loss (SSNHL) may have word recognition scores (WRS) that correlate with pure tone average (PTA). We hypothesize that there is a subset of patients with SSNHL who have improved WRS despite stable PTA. Methods:Retrospective case review at a tertiary otolaryngology practice. Results:We identified 13 of 113 patients with SSNHL whose WRS increased despite overall sta-ble pure tone averages. There was an observed average improvement in WRS by 23.8 points in this patient cohort at follow-up, with mean initial PTA in the affected ear at 48.7 dB. Conclusions: We identify a novel cohort of SSNHL patients that have failed treatment as measured by PTA, but who have increased WRS over time. These data have implications for patient counseling and lend insight into the pathophysiology of SSNHL.

A Discussion of Foreign Language Listening Problems and Their Causes among Intermediate EFL learners in Chinese universities

Listening has been considered as the most important language skill by many researchers but in Chinese context it has been overlooked for years.The Foreign Language(FL)listening problems gradually reveal when students attend to the Listening Course in universities or colleges.This paper at first illustrates the importance of listening in foreign language learning and provides a discussion on the top-down and bottom-up processes with the analysis of Anderson's three-model of language comprehension[9]as well as the factors affecting listening comprehension in Foreign Language in general,aiming at the illustration of the necessity on one of the bottom-up processes-the Spoken Word Recognition(SWR).The analysis of two issued official guiding documents for Chinese university students and the current situation of English listening teaching in Chinese universities are followed,providing a general background on English listening teaching in Chinese context.Three possible foreign language listening problems and their causes are proposed,followed by the pedagogical suggestions to EFL learners who are improving their listening skills and instructors who are advancing their teaching procedures.

Effect of face masks on speech understanding: A clinical perspective during speech audiometry

The objective was to measure the effect of various face masks on speech recognition threshold and the word recognition score in the presence of varying background noise levels.20 normal-hearing adult subjects(a total of 40 ears)participated.Pure tone audiometry followed by speech recognition threshold and word recognition score at the most comfortable level in varying signal-to-noise ratios(SNR0,SNR10,and SNR15)using surgical,pleated cloth,and N95 masks.Using surgical,cloth,and N95 masks,speech recognition thresholds increased by 1.8 dB,4.4 dB,and 5.05 dB,respectively.Word recognition scores decreased by 32%without a mask,43.7%in a surgical mask,46.3%in a cloth mask,and 46.7%in N95 mask conditions,between SNR15 and SNR0.The speech recognition threshold was negatively affected with cloth and N95 masks.Surgical masks do not affect the word recognition scores at lower background noise levels.However,as the signal-to-noise ratio decreased,even the surgical,cloth,and N95 masks significantly impacted the word recognition score even in normal-hearing individuals.

Hearing Impairment in Senile Dementia of Alzhaimer's Type

Objectives To evaluate peripheral auditory dysfunction in senile dementia of Alzheimer’s disease (AD) and its relationship with cognitive dysfunction. Methods Pure tone thresholds,word recognition scores (WRS), acoustic immittance and auditory brain-stem responses (ABR) were tested to evaluate the auditory function in 43 AD patients and 50 normal subjects. The test reliability in these subjects was examined before the test results were evaluated for their correlation with the Mini Mental State Examination(MMSE) score. Results There were no statistically significant differences in peripheral auditory functions between the two ears in the tested subjects or between the two groups when the auditometric results of the right ear were compared(P > 0.05). Also, there were no statistically significant differences between the two groups when audiometric test reliability, acoustic impedance and ABR results were compared(P > 0.05). Conclutions The pure tone audiometric threshold and WRS in AD patients are similar to those in comparable non-AD senile subjects. Peripheral auditory dysfunction is not related to cognitive dysfunction.

Effectiveness of Context-Based Strategy for English Learning among Chinese Middle School Students

Currently,a growing number of educators are aware of the need to look for new approaches to replace the"spoon-feeding"method.Therefore,the context-based strategy began to emerge.This study aims to investigate how students derive information through contextual clues by examining the progress of Chinese middle school EFL students in terms of word recognition.The participants of this study included 20 eighth grade students from the same middle school.The participants sat for two different quizzes:a contextual vocabulary quiz(quiz A)and a direct instruction quiz(quiz B).In quiz A,the participants inferred the meaning of the target words from the example sentences,whereas in quiz B,the students utilized the accompanying English explanation to guess other new words.These students were in the experimental and control conditions,respectively.The two quizzes comprised of 15 multiple choice questions(MCQ)which differentiated the participants?word recognition response to two different learning methods.There were two significant findings from this study.First,the results showed that the context-based strategy leads to a better vocabulary learning performance compared to the direct instruction strategy.Second,although it is not as effective as the context-based strategy,the direct instruction strategy may assist EFL learners in remembering words in short term.

Development of SC-10: A psychometrically equivalent Singapore Mandarin disyllabic word list for clinical speech audiometry use

Objective::The aim of this study is to develop and evaluate a set of psychometrically equivalent disyllabic wordlist(SC-10)in Singapore Mandarin for clinical use.Study design::A preliminary set of 1000 words were obtained from a dictionary of frequently used words by Singapore students.Ten native judges rate the familiarity level of each word.This is followed by a face-to-face public survey to rank the shortlisted word set from most to least familiar.The final 108 disyllabic words were recorded by a native female talker.20 normal hearing subjects were used to obtain the percentage of correct word recognition at 24 intensity levels(-10 dB HL to 26 dB HL in 2 dB increment).Psychometric function slopes were calculated for each word.100 words were eventually chosen and assigned into ten 10-word lists based on a psychometric balancing protocol.Minor digital adjustments were made to the intensity of each wordlist to improve their auditory homogeneity.Results::The calculated regression slopes in the psychometric functions for the ten lists are between 8.0 and 9.8%/dB.Single factor ANOVA analysis showed no significant difference in both the mean intensity required to obtain 50%recognition score(f=0.109,df=9,p=0.999)and the slopes of the psychometric functions(f=0.078,df=9,p=0.999)between the ten word lists.List validation on 25 normal hearing participants(PTA=11.0 dB HL,SD=4.3)showed a mean speech recognition threshold(SRT)of 9.3 dB HL(SD=3.5)and regression slope of 8.395%/dB.Quadratic regression analysis showed a positive correlation(r^(2)=0.923)between presentation level and word recognition score(WRS).The difference between PTA and SRT of each subject all fall within the clinically acceptable difference of 10 dB HL.Test-retest reliability,carried out on 20 subjects at three levels(SRT-5,SRT,and SRT+5 dB),showed no significance difference between word recognition score when the same participant is tested again at the same intensity level using a different wordlist.Conclusion::All in all,it shows that the SC-10 speech materials are valid for clinical use for Mandarin speech audiometry in Singapore.

WHAT DO CHINESE EFL LISTENERS NEED TO IMPROVE: BOTTOM-UP SKILLS OR TOP-DOWN STRATEGIES?— A COGNITIVE PERSPECTIVE ON CHINESE EFL LISTENERS' ROOT PROBLEMS

L2 listening comprehension is a cognitive process, in which listeners use both bottom-up and top-down processing to comprehend the aural text. This study explores whether Chinese EFL learners' listening problems are more associated with their bottom-up or top-down processing. A questionnaire and students' verbal report were conducted. The results show that both high-and low-proficient listeners are actively engaged in top-down processing, but their degree of comprehension depends, to a large extent, on their success in bottom-up processing. This study appeals for more focus shifted back to training Chinese EFL learners' bottom-up processing ability.

The Brain Connectome for Chinese Reading

Chinese,as a logographic language,fundamentally differs from alphabetic languages like English.Previous neuroimaging studies have mainly focused on alphabetic languages,while the exploration of Chinese reading is still an emerging and fast-growing research field.Recently,a growing number of neuroimaging studies have explored the neural circuit of Chinese reading.Here,we summarize previous research on Chinese reading from a connectomic perspective.Converging evidence indicates that the left middle frontal gyrus is a specialized hub region that connects the ventral with dorsal pathways for Chinese reading.Notably,the orthography-to-phonology and orthography-to-semantics mapping,mainly processed in the ventral pathway,are more specific during Chinese reading.Besides,in addition to the left-lateralized language-related regions,reading pathways in the right hemisphere also play an important role in Chinese reading.Throughout,we comprehensively review prior findings and emphasize several challenging issues to be explored in future work.

Sequential Back-Propagation

In this paper we consider the problem of sequential processing and present a sequen-tial model based on the back-propagation algorithm. This model is intended to deal with intrinsically sequential problems, such as word recognition, speech recognition,natural language understanding. This model can be used to train a network to learn the sequence of input patterns, in a fixed order or a random order. Besides, this mod-el is open- and partial-associative, characterized as 'recognizing while accumulating',which, as we argue, is mental cognition process oriented.