Search results for: speech recognition performance

Authors: Zoran Spasovski

Abstract:

The paper shortly discusses the significance and the benefits of the historical approach in the research of languages by presenting examples of it in the fields of phonetics and phonology, lexicology, morphology, syntax, and even in the onomastics (toponomy and anthroponomy). The examples from the field of phonetics/phonology include insights into animal speech and its evolution into human speech, the evolution of the sounds of human speech from vocals to glides and consonants and from velar consonants to palatal, etc., on well-known examples of former researchers. Those from the field of lexicology show shortly the formation of the lexemes and their evolution; the morphology and syntax are explained by examples of the development of grammar and syntax forms, and the importance of the historical approach in the research of place-names and personal names is briefly outlined through examples of place-names and personal names and surnames, and the conclusions that come from it, in different languages.

Keywords: animal speech, glotogenesis, grammar forms, lexicology, place-names, personal names, surnames, syntax categories

Procedia PDF Downloads 83

Authors: Kilari Nikhil, Ankur Tibrewal, Srinivas Kruthiventi S. S.

Abstract:

Conventional speaker identification systems often fall short of capturing the diverse variations present in speech data due to fixed-scale architectures. In this research, we propose a CNN-based architecture, USENet, designed to overcome these limitations. Leveraging two key techniques, our approach achieves superior performance on the VoxCeleb 1 Dataset without any pre-training. Firstly, we adopt a U-net-inspired design to extract features at multiple scales, empowering our model to capture speech characteristics effectively. Secondly, we introduce the squeeze and excitation block to enhance spatial feature learning. The proposed architecture showcases significant advancements in speaker identification, outperforming existing methods, and holds promise for future research in this domain.

Keywords: multi-scale feature extraction, squeeze and excitation, VoxCeleb1 speaker identification, mel-spectrograms, USENet

Procedia PDF Downloads 74

Authors: Ana Londral, Burcu Demiray, Marcus Cheetham

Abstract:

Speech recording is a methodology used in many different studies related to cognitive and behaviour research. Modern advances in digital equipment brought the possibility of continuously recording hours of speech in naturalistic environments and building rich sets of sound files. Speech analysis can then extract from these files multiple features for different scopes of research in Language and Communication. However, tools for analysing a large set of sound files and automatically extract relevant features from these files are often inaccessible to researchers that are not familiar with programming languages. Manual analysis is a common alternative, with a high time and efficiency cost. In the analysis of long sound files, the first step is the voice segmentation, i.e. to detect and label segments containing speech. We present a comprehensive methodology aiming to support researchers on voice segmentation, as the first step for data analysis of a big set of sound files. Praat, an open source software, is suggested as a tool to run a voice detection algorithm, label segments and files and extract other quantitative features on a structure of folders containing a large number of sound files. We present the validation of our methodology with a set of 5000 sound files that were collected in the daily life of a group of voluntary participants with age over 65. A smartphone device was used to collect sound using the Electronically Activated Recorder (EAR): an app programmed to record 30-second sound samples that were randomly distributed throughout the day. Results demonstrated that automatic segmentation and labelling of files containing speech segments was 74% faster when compared to a manual analysis performed with two independent coders. Furthermore, the methodology presented allows manual adjustments of voiced segments with visualisation of the sound signal and the automatic extraction of quantitative information on speech. In conclusion, we propose a comprehensive methodology for voice segmentation, to be used by researchers that have to work with large sets of sound files and are not familiar with programming tools.

Keywords: automatic speech analysis, behavior analysis, naturalistic environments, voice segmentation

Procedia PDF Downloads 281

Authors: Amulya P. Rao, Prathima S., Sreedevi N.

Abstract:

Speech sound disorders (SSD) encompass the major concern in younger population of India with highest prevalence rate among the speech disorders. Children with SSD if not identified and rehabilitated at the earliest, are at risk for academic difficulties. This necessitates early identification using screening tools assessing the frequently misarticulated speech sounds. The literature on frequently misarticulated speech sounds is ample in English and other western languages targeting individuals with various communication disorders. Articulation is language specific, and there are limited studies reporting the same in Kannada, a Dravidian Language. Hence, the present study aimed to identify the frequently misarticulated consonants in Kannada and also to examine the error type. A retrospective, descriptive study was carried out using secondary data analysis of 41 participants (34-phonetic type and 7-phonemic type) with SSD in the age range 3-to 12-years. All the consonants of Kannada were analyzed by considering three words for each speech sound from the Kannada Diagnostic Photo Articulation test (KDPAT). Picture naming task was carried out, and responses were audio recorded. The recorded data were transcribed using IPA 2018 broad transcription. A criterion of 2/3 or 3/3 error productions was set to consider the speech sound to be an error. Number of error productions was calculated for each consonant in each participant. Then, the percentage of participants meeting the criteria were documented for each consonant to identify the frequently misarticulated speech sound. Overall results indicated that velar /k/ (48.78%) and /g/ (43.90%) were frequently misarticulated followed by voiced retroflex /ɖ/ (36.58%) and trill /r/ (36.58%). The lateral retroflex /ɭ/ was misarticulated by 31.70% of the children with SSD. Dentals (/t/, /n/), bilabials (/p/, /b/, /m/) and labiodental /v/ were produced correctly by all the participants. The highly misarticulated velars /k/ and /g/ were frequently substituted by dentals /t/ and /d/ respectively or omitted. Participants with SSD-phonemic type had multiple substitutions for one speech sound whereas, SSD-phonetic type had consistent single sound substitutions. Intra- and inter-judge reliability for 10% of the data using Cronbach’s Alpha revealed good reliability (0.8 ≤ α < 0.9). Analyzing a larger sample by replicating such studies will validate the present study results.

Keywords: consonant, frequently misarticulated, Kannada, SSD

Procedia PDF Downloads 134

Authors: Hamsini Pulugurtha, P. V. S. L. Jagadamba

Abstract:

Programmed Speech to Text and Text Summarization Using Graph-based Algorithms can be utilized in gatherings to get the short depiction of the gathering for future reference. This gives signature check utilizing Siamese neural organization to confirm the personality of the client and convert the client gave sound record which is in English into English text utilizing the discourse acknowledgment bundle given in python. At times just the outline of the gathering is required, the answer for this text rundown. Thus, the record is then summed up utilizing the regular language preparing approaches, for example, solo extractive text outline calculations

Keywords: Siamese neural network, English speech, English text, natural language processing, unsupervised extractive text summarization

Procedia PDF Downloads 218

Authors: Qin Long, Li Xiaoge

Abstract:

The Knowledge Graphs have now become a new form of knowledge representation. However, there is no consensus in regard to a plausible and definition of entities and relationships in the domain-specific knowledge graph. Further, in conjunction with several limitations and deficiencies, various domain-specific entities and relationships recognition approaches are far from perfect. Specifically, named entity recognition in Chinese domain is a critical task for the natural language process applications. However, a bottleneck problem with Chinese named entity recognition in new domains is the lack of annotated data. To address this challenge, a domain distant supervised named entity recognition framework is proposed. The framework is divided into two stages: first, the distant supervised corpus is generated based on the entity linking model of graph attention neural network; secondly, the generated corpus is trained as the input of the distant supervised named entity recognition model to train to obtain named entities. The link model is verified in the ccks2019 entity link corpus, and the F1 value is 2% higher than that of the benchmark method. The re-pre-trained BERT language model is added to the benchmark method, and the results show that it is more suitable for distant supervised named entity recognition tasks. Finally, it is applied in the computer field, and the results show that this framework can obtain domain named entities.

Keywords: distant named entity recognition, entity linking, knowledge graph, graph attention neural network

Procedia PDF Downloads 93

Authors: Reza Shakoori

Abstract:

This paper presents a structural correction of Arabic and Persian strokes using manipulation of their chain codes in order to improve the rate and performance of Persian and Arabic handwritten word recognition systems. It collects pure and effective features to represent a character with one consolidated feature vector and reduces variations in order to decrease the number of training samples and increase the chance of successful classification. Our results also show that how the proposed approaches can simplify classification and consequently recognition by reducing variations and possible noises on the chain code by keeping orientation of characters and their backbone structures.

Keywords: Arabic, chain code normalization, OCR systems, image processing

Procedia PDF Downloads 404

Authors: Jonathan Sohn

Abstract:

Deep neural networks (DNNs) have shown state-of-the-art performance for many applications, including computer vision, natural language processing, and speech recognition. Recently, adversarial attacks have been studied in the context of deep neural networks, which aim to alter the results of deep neural networks by modifying the inputs slightly. For example, an adversarial attack on a DNN used for object detection can cause the DNN to miss certain objects. As a result, the reliability of DNNs is undermined by their lack of robustness against adversarial attacks, raising concerns about their use in safety-critical applications such as autonomous driving. In this paper, we focus on studying the adversarial attacks and defenses on DNNs for image classification. There are two types of adversarial attacks studied which are fast gradient sign method (FGSM) attack and projected gradient descent (PGD) attack. A DNN forms decision boundaries that separate the input images into different categories. The adversarial attack slightly alters the image to move over the decision boundary, causing the DNN to misclassify the image. FGSM attack obtains the gradient with respect to the image and updates the image once based on the gradients to cross the decision boundary. PGD attack, instead of taking one big step, repeatedly modifies the input image with multiple small steps. There is also another type of attack called the target attack. This adversarial attack is designed to make the machine classify an image to a class chosen by the attacker. We can defend against adversarial attacks by incorporating adversarial examples in training. Specifically, instead of training the neural network with clean examples, we can explicitly let the neural network learn from the adversarial examples. In our experiments, the digit recognition accuracy on the MNIST dataset drops from 97.81% to 39.50% and 34.01% when the DNN is attacked by FGSM and PGD attacks, respectively. If we utilize FGSM training as a defense method, the classification accuracy greatly improves from 39.50% to 92.31% for FGSM attacks and from 34.01% to 75.63% for PGD attacks. To further improve the classification accuracy under adversarial attacks, we can also use a stronger PGD training method. PGD training improves the accuracy by 2.7% under FGSM attacks and 18.4% under PGD attacks over FGSM training. It is worth mentioning that both FGSM and PGD training do not affect the accuracy of clean images. In summary, we find that PGD attacks can greatly degrade the performance of DNNs, and PGD training is a very effective way to defend against such attacks. PGD attacks and defence are overall significantly more effective than FGSM methods.

Keywords: deep neural network, adversarial attack, adversarial defense, adversarial machine learning

Procedia PDF Downloads 195

Authors: Zsuzsanna Schnell, Francesca Ervas

Abstract:

Background: the experimental pragmatic study confirms a basic tenet in the Relevance theoretical views in language philosophy. It draws up a developmental trajectory of the maxims, revealing the cognitive difficulty of their interpretation, their relative place to each other, and the order they may follow in development. A central claim of the present research is that social-cognitive skills play a significant role in inferential meaning construction. Children passing the False Belief Test are significantly more successful in tasks measuring the recognition of the infringement of conversational maxims. Aims and method: Preschoolers’ conversational skills and pragmatic competence is examined in view of their mentalization skills. In doing so it use a measure of linguistic tasks, containing 5 short scenarios for each Gricean maxim. it measure preschoolers’ ToM performance with a first- and a second order ToM task and compare participants’ ability to recognize the infringement of the Gricean maxims in view of their social cognitive skills. Results: Findings suggest that Theory of Mind has a predictive force of 75% concerning the ability to follow Gricean maxims efficiently. ToM proved to be a significant factor in predicting the group’s performance and success rates in 3 out of 4 maxim infringement recognition tasks: in the Quantity, Relevance and Manner conditions, but not in the Quality trial. Conclusions: the results confirm that children’s communicative competence in social contexts requires the development of higher-order social-cognitive reasoning, and reveal the cognitive effort needed for the recognition of the infringement of each maxim, yielding a continuum of their cognitive difficulty and trajectory of development.

Keywords: maxim infringement recognition, social cognition, Gricean maxims, developmental pragmatics

Procedia PDF Downloads 7

Authors: Shrabani Ghosh

Abstract:

Automatic offensive language detection in social media has become a stirring task in today's NLP. Manual Offensive language detection is tedious and laborious work where automatic methods based on machine learning are only alternatives. Previous works have done sentiment analysis over social media in different ways such as supervised, semi-supervised, and unsupervised manner. Domain adaptation in a semi-supervised way has also been explored in NLP, where the source domain and the target domain are different. In domain adaptation, the source domain usually has a large amount of labeled data, while only a limited amount of labeled data is available in the target domain. Pretrained transformers like BERT, RoBERTa models are fine-tuned to perform text classification in an unsupervised manner to perform further pre-train masked language modeling (MLM) tasks. In previous work, hate speech detection has been explored in Gab.ai, which is a free speech platform described as a platform of extremist in varying degrees in online social media. In domain adaptation process, Twitter data is used as the source domain, and Gab data is used as the target domain. The performance of domain adaptation also depends on the cross-domain similarity. Different distance measure methods such as L2 distance, cosine distance, Maximum Mean Discrepancy (MMD), Fisher Linear Discriminant (FLD), and CORAL have been used to estimate domain similarity. Certainly, in-domain distances are small, and between-domain distances are expected to be large. The previous work finding shows that pretrain masked language model (MLM) fine-tuned with a mixture of posts of source and target domain gives higher accuracy. However, in-domain performance of the hate classifier on Twitter data accuracy is 71.78%, and out-of-domain performance of the hate classifier on Gab data goes down to 56.53%. Recently self-supervised learning got a lot of attention as it is more applicable when labeled data are scarce. Few works have already been explored to apply self-supervised learning on NLP tasks such as sentiment classification. Self-supervised language representation model ALBERTA focuses on modeling inter-sentence coherence and helps downstream tasks with multi-sentence inputs. Self-supervised attention learning approach shows better performance as it exploits extracted context word in the training process. In this work, a self-supervised attention mechanism has been proposed to detect hate speech on Gab.ai. This framework initially classifies the Gab dataset in an attention-based self-supervised manner. On the next step, a semi-supervised classifier trained on the combination of labeled data from the first step and unlabeled data. The performance of the proposed framework will be compared with the results described earlier and also with optimized outcomes obtained from different optimization techniques.

Keywords: attention learning, language model, offensive language detection, self-supervised learning

Procedia PDF Downloads 105

Authors: Lanchi Xie, Zhihui Li, Zhigang Li, Guiqiang Wang, Lei Xu, Yuwen Yan

Abstract:

Image-based facial features can be classified into category recognition features and individual recognition features. Current automated face recognition systems extract a specific feature vector of different dimensions from a facial image according to their pre-trained neural network. However, to improve the efficiency of parameter calculation, an algorithm generally reduces the image details by pooling. The operation will overlook the details concerned much by forensic experts. In our experiment, we adopted a variety of face recognition algorithms based on deep learning, compared a large number of naturally collected face images with the known data of the same person's frontal ID photos. Downscaling and manual handling were performed on the testing images. The results supported that the facial recognition algorithms based on deep learning detected structural and morphological information and rarely focused on specific markers such as stains and moles. Overall performance, distribution of genuine scores and impostor scores, and likelihood ratios were tested to evaluate the accuracy of biometric systems and forensic experts. Experiments showed that the biometric systems were skilled in distinguishing category features, and forensic experts were better at discovering the individual features of human faces. In the proposed approach, a fusion was performed at the score level. At the specified false accept rate, the framework achieved a lower false reject rate. This paper contributes to improving the interpretability of the objective method of facial comparison and provides a novel method for human-machine collaboration in this field.

Keywords: likelihood ratio, automated facial recognition, facial comparison, biometrics

Procedia PDF Downloads 130

Authors: Teodora Vukovic, Christoph Hottiger, Noah Bubenhofer

Abstract:

The development of VIAN-DH aims at bridging two linguistic approaches: conversation analysis/interactional linguistics (IL), so far a dominantly qualitative field, and computational/corpus linguistics and its quantitative and automated methods. Contemporary IL investigates the systematic organization of conversations and interactions composed of speech, gaze, gestures, and body positioning, among others. These highly integrated multimodal behaviour is analysed based on video data aimed at uncovering so called “multimodal gestalts”, patterns of linguistic and embodied conduct that reoccur in specific sequential positions employed for specific purposes. Multimodal analyses (and other disciplines using videos) are so far dependent on time and resource intensive processes of manual transcription of each component from video materials. Automating these tasks requires advanced programming skills, which is often not in the scope of IL. Moreover, the use of different tools makes the integration and analysis of different formats challenging. Consequently, IL research often deals with relatively small samples of annotated data which are suitable for qualitative analysis but not enough for making generalized empirical claims derived quantitatively. VIAN-DH aims to create a workspace where many annotation layers required for the multimodal analysis of videos can be created, processed, and correlated in one platform. VIAN-DH will provide a graphical interface that operates state-of-the-art tools for automating parts of the data processing. The integration of tools that already exist in computational linguistics and computer vision, facilitates data processing for researchers lacking programming skills, speeds up the overall research process, and enables the processing of large amounts of data. The main features to be introduced are automatic speech recognition for the transcription of language, automatic image recognition for extraction of gestures and other visual cues, as well as grammatical annotation for adding morphological and syntactic information to the verbal content. In the ongoing instance of VIAN-DH, we focus on gesture extraction (pointing gestures, in particular), making use of existing models created for sign language and adapting them for this specific purpose. In order to view and search the data, VIAN-DH will provide a unified format and enable the import of the main existing formats of annotated video data and the export to other formats used in the field, while integrating different data source formats in a way that they can be combined in research. VIAN-DH will adapt querying methods from corpus linguistics to enable parallel search of many annotation levels, combining token-level and chronological search for various types of data. VIAN-DH strives to bring crucial and potentially revolutionary innovation to the field of IL, (that can also extend to other fields using video materials). It will allow the processing of large amounts of data automatically and, the implementation of quantitative analyses, combining it with the qualitative approach. It will facilitate the investigation of correlations between linguistic patterns (lexical or grammatical) with conversational aspects (turn-taking or gestures). Users will be able to automatically transcribe and annotate visual, spoken and grammatical information from videos, and to correlate those different levels and perform queries and analyses.

Keywords: multimodal analysis, corpus linguistics, computational linguistics, image recognition, speech recognition

Procedia PDF Downloads 108

Authors: Pavan K. Rallabandi, Kailash C. Patidar

Abstract:

In this paper, we present an optimization technique or a learning algorithm using the hybrid architecture by combining the most popular sequence recognition models such as Recurrent Neural Networks (RNNs) and Hidden Markov models (HMMs). In order to improve the sequence or pattern recognition/ classification performance by applying a hybrid/neural symbolic approach, a gradient descent learning algorithm is developed using the Real Time Recurrent Learning of Recurrent Neural Network for processing the knowledge represented in trained Hidden Markov Models. The developed hybrid algorithm is implemented on automata theory as a sample test beds and the performance of the designed algorithm is demonstrated and evaluated on learning the deterministic finite state automata.

Keywords: hybrid systems, hidden markov models, recurrent neural networks, deterministic finite state automata

Procedia PDF Downloads 388

Authors: Stefania Arguelles Reyes, Octavio José Salcedo Parra, Alberto Acosta López

Abstract:

This paper presents a web application for the improvement of images through recognition. The web application is based on the analysis of picture-based recognition methods that allow an improvement on the physical appearance of people posting in social networks. The basis relies on the study of tools that can correct or improve some features of the face, with the help of a wide collection of user images taken as reference to build a facial profile. Automatic facial profiling can be achieved with a deeper study of the Object Detection Library. It was possible to improve the initial images with the help of MATLAB and its filtering functions. The user can have a direct interaction with the program and manually adjust his preferences.

Keywords: Matlab, make up, recognition methods, web application

Procedia PDF Downloads 144

Authors: Andre Wittenborn, Jarek Krajewski

Abstract:

Trauma-related sadness in speech can alter the voice in several ways. The generation of non-linear aerodynamic phenomena within the vocal tract is crucial when analyzing trauma-influenced speech production. They include non-laminar flow and formation of jets rather than well-behaved laminar flow aspects. Especially state-space reconstruction methods based on chaotic dynamics and fractal theory have been suggested to describe these aerodynamic turbulence-related phenomena of the speech production system. To extract the non-linear properties of the speech signal, we used the time delay embedding method to reconstruct from a scalar time series (reconstructed phase space, RPS). This approach results in the extraction of 7238 Features per .wav file (N= 47, 32 m, 15 f). The speech material was prompted by telling about autobiographical related sadness-inducing experiences (sampling rate 16 kHz, 8-bit resolution). After combining these features in a support vector machine based machine learning approach (leave-one-sample out validation), we achieved a correlation of r = .41 with the well-established, self-report ground truth measure (RATS) of post-traumatic stress disorder (PTSD).

Keywords: non-linear dynamics features, post traumatic stress disorder, reconstructed phase space, support vector machine

Procedia PDF Downloads 102

Authors: Frederik Calsius, Mirela Popa, Alexia Briassouli

Abstract:

In the field of machine learning, it is common practice to use benchmark datasets to prove the working of a method. The domain of action recognition in videos often uses datasets like Kinet-ics, Something-Something, UCF-101 and HMDB-51 to report results. Considering the properties of the datasets, there are no datasets that focus solely on very short clips (2 to 3 seconds), and on highly-similar fine-grained actions within one specific domain. This paper researches how current state-of-the-art action recognition methods perform on a dataset that consists of highly similar, fine-grained actions. To do so, a dataset of skateboarding tricks was created. The performed analysis highlights both benefits and limitations of state-of-the-art methods, while proposing future research directions in the activity recognition domain. The conducted research shows that the best results are obtained by fusing RGB data with OpenPose data for the Temporal Shift Module.

Keywords: activity recognition, fused deep representations, fine-grained dataset, temporal modeling

Procedia PDF Downloads 231

Authors: M. Pilgun

Abstract:

The report presents the results of a study of the specifics of speech perception by actors of video hosting services on the material of urban planning conflicts. To analyze the content, the multimodal approach using neural network technologies is employed. Analysis of word associations and associative networks of relevant stimulus revealed the evaluative reactions of the actors. Analysis of the data identified key topics that generated negative and positive perceptions from the participants. The calculation of social stress and social well-being indices based on user-generated content made it possible to build a rating of road transport construction objects according to the degree of negative and positive perception by actors.

Keywords: social media, speech perception, video hosting, networks

Procedia PDF Downloads 147

Authors: Natalia V. Ursul

Abstract:

In linguistic studies, the question of nonsense is attracting increasing interest. Nonsense is usually defined as spoken or written words that have no meaning. However, this definition is likely to be outdated as any speech act is generated due to the speaker’s pragmatic reasons, thus it cannot be purely illogical or meaningless. In the current paper a new working definition of nonsense as a linguistic medium will be formulated; moreover, the pragmatic peculiarities of newly coined linguistic patterns and possible ways of their interpretation will be discussed.

Keywords: nonsense, nonse verse, pragmatics, speech act

Procedia PDF Downloads 519

Authors: Tsvetomira Braynova, Miglena Simonska

Abstract:

The article presents the results of research on phonological processes in three and four-year-old children. For the purpose of the study, an author's test was developed and conducted among 120 children. The study included three areas of research - at the level of words (96 words), at the level of sentence repetition (10 sentences) and at the level of generating own speech from a picture (15 pictures). The test also gives us additional information about the articulation errors of the assessed children. The main purpose of the icing is to analyze all phonological processes that occur at this age in Bulgarian children and to identify which are typical and atypical for this age. The results show that the most common phonology errors that children make are: sound substitution, an elision of sound, metathesis of sound, elision of a syllable, and elision of consonants clustered in a syllable. All examined children were identified with the articulatory disorder from type bilabial lambdacism. Measuring the correlation between the average length of repeated speech and the average length of generated speech, the analysis proves that the more words a child can repeat in part “repeated speech,” the more words they can be expected to generate in part “generating sentence.” The results of this study show that the task of naming a word provides sufficient and representative information to assess the child's phonology.

Keywords: assessment, phonology, articulation, speech-language development

Procedia PDF Downloads 186

Authors: Rodolfo Lorbieski, Silvia Modesto Nassar

Abstract:

Many supervised machine learning tasks require decision making across numerous different classes. Multi-class classification has several applications, such as face recognition, text recognition and medical diagnostics. The objective of this article is to analyze an adapted method of Stacking in multi-class problems, which combines ensembles within the ensemble itself. For this purpose, a training similar to Stacking was used, but with three levels, where the final decision-maker (level 2) performs its training by combining outputs from the tree-based pair of meta-classifiers (level 1) from Bayesian families. These are in turn trained by pairs of base classifiers (level 0) of the same family. This strategy seeks to promote diversity among the ensembles forming the meta-classifier level 2. Three performance measures were used: (1) accuracy, (2) area under the ROC curve, and (3) time for three factors: (a) datasets, (b) experiments and (c) levels. To compare the factors, ANOVA three-way test was executed for each performance measure, considering 5 datasets by 25 experiments by 3 levels. A triple interaction between factors was observed only in time. The accuracy and area under the ROC curve presented similar results, showing a double interaction between level and experiment, as well as for the dataset factor. It was concluded that level 2 had an average performance above the other levels and that the proposed method is especially efficient for multi-class problems when compared to binary problems.

Keywords: stacking, multi-layers, ensemble, multi-class

Procedia PDF Downloads 269

Authors: Kyi Pyar Zaw, Zin Mar Kyu

Abstract:

Character recognition is the process of converting a text image file into editable and searchable text file. Feature Extraction is the heart of any character recognition system. The character recognition rate may be low or high depending on the extracted features. In the proposed paper, 25 features for one character are used in character recognition. Basically, there are three steps of character recognition such as character segmentation, feature extraction and classification. In segmentation step, horizontal cropping method is used for line segmentation and vertical cropping method is used for character segmentation. In the Feature extraction step, features are extracted in two ways. The first way is that the 8 features are extracted from the entire input character using eight direction chain code frequency extraction. The second way is that the input character is divided into 16 blocks. For each block, although 8 feature values are obtained through eight-direction chain code frequency extraction method, we define the sum of these 8 feature values as a feature for one block. Therefore, 16 features are extracted from that 16 blocks in the second way. We use the number of holes feature to cluster the similar characters. We can recognize the almost Myanmar common characters with various font sizes by using these features. All these 25 features are used in both training part and testing part. In the classification step, the characters are classified by matching the all features of input character with already trained features of characters.

Keywords: chain code frequency, character recognition, feature extraction, features matching, segmentation

Procedia PDF Downloads 320

Authors: Si Chen, Quanhong Jiang

Abstract:

In the increasingly mature posture recognition technology, human movement information is widely used in sports rehabilitation, human-computer interaction, medical health, human posture assessment, and other fields today; this project uses the most original ideas; it is proposed to use the collection equipment for the collection of myoelectric data, reflect the muscle posture change on a degree of freedom through data processing, carry out data-muscle three-dimensional model joint adjustment, and realize basic pose recognition. Based on this, bionic aids or medical rehabilitation equipment can be further developed with the help of robotic arms and cutting-edge technology, which has a bright future and unlimited development space.

Keywords: pose recognition, 3D animation, electromyography, machine learning, bionics

Procedia PDF Downloads 79

Authors: Aristi Alopoudi, Sofia Beloka, Vassiliki Pliogou

Abstract:

Autism is a complex neuro-developmental disorder with a variety of difficulties in many aspects such as social interaction, communication skills and verbal communication (speech). The aim of this study was to examine the impact of therapeutic horseback riding in improving the verbal and communication skills of children diagnosed with autism during 16 sessions. The researcher examined whether the expression of speech, the use of vocabulary, semantics, pragmatics, echolalia and communication skills were influenced by the therapeutic horseback riding when we increase the frequency of the sessions. The researcher observed two subjects of primary-school aged, in a two case observation design, with autism during 16 therapeutic horseback riding sessions (one riding session per week). Compared to baseline, at the end of the 16th therapeutic session, therapeutic horseback riding increased both verbal skills such as vocabulary, semantics, pragmatics, formation of sentences and communication skills such as eye contact, greeting, participation in dialogue and spontaneous speech. It was noticeable that echolalia remained stable. Increased frequency of therapeutic horseback riding was beneficial for significant improvement in verbal and communication skills. More specifically, from the first to the last riding session there was a great increase of vocabulary, semantics, and formation of sentences. Pragmatics reached a lower level than semantics but the same as the right usage of the first person (for example, I make a hug) and echolalia used for that. A great increase of spontaneous speech was noticed. The eye contact was presented in a lower level, and there was a slow but important raise at the greeting as well as the participation in dialogue. Last but not least; this is a first study conducted in therapeutic horseback riding studying the verbal communication and communication skills in autistic children. According to the references, therapeutic horseback riding is a therapy with a variety of benefits, thus; this research made clear that in the benefits of this therapy there should be included the improvement of verbal speech and communication.

Keywords: Autism, communication skills, speech, therapeutic horseback riding

Procedia PDF Downloads 274

Authors: Yanting Cao, Kazumitsu Nawata

Abstract:

As smartphones upgrading, their software and hardware are getting smarter, so the smartphone-based human activity recognition will be described as more refined, complex, and detailed. In this context, we analyzed a set of experimental data obtained by observing and measuring 30 volunteers with six activities of daily living (ADL). Due to the large sample size, especially a 561-feature vector with time and frequency domain variables, cleaning these intractable features and training a proper model becomes extremely challenging. After a series of feature selection and parameters adjustment, a well-performed SVM classifier has been trained.

Keywords: smart sensors, human activity recognition, artificial intelligence, SVM

Procedia PDF Downloads 144

Authors: Ben Soltane Cheima, Ittansa Yonas Kelbesa

Abstract:

Speaker Identification (SI) is the task of establishing identity of an individual based on his/her voice characteristics. The SI task is typically achieved by two-stage signal processing: training and testing. The training process calculates speaker specific feature parameters from the speech and generates speaker models accordingly. In the testing phase, speech samples from unknown speakers are compared with the models and classified. Even though performance of speaker identification systems has improved due to recent advances in speech processing techniques, there is still need of improvement. In this paper, a Closed-Set Tex-Independent Speaker Identification System (CISI) based on a Multiple Classifier System (MCS) is proposed, using Mel Frequency Cepstrum Coefficient (MFCC) as feature extraction and suitable combination of vector quantization (VQ) and Gaussian Mixture Model (GMM) together with Expectation Maximization algorithm (EM) for speaker modeling. The use of Voice Activity Detector (VAD) with a hybrid approach based on Short Time Energy (STE) and Statistical Modeling of Background Noise in the pre-processing step of the feature extraction yields a better and more robust automatic speaker identification system. Also investigation of Linde-Buzo-Gray (LBG) clustering algorithm for initialization of GMM, for estimating the underlying parameters, in the EM step improved the convergence rate and systems performance. It also uses relative index as confidence measures in case of contradiction in identification process by GMM and VQ as well. Simulation results carried out on voxforge.org speech database using MATLAB highlight the efficacy of the proposed method compared to earlier work.

Keywords: feature extraction, speaker modeling, feature matching, Mel frequency cepstrum coefficient (MFCC), Gaussian mixture model (GMM), vector quantization (VQ), Linde-Buzo-Gray (LBG), expectation maximization (EM), pre-processing, voice activity detection (VAD), short time energy (STE), background noise statistical modeling, closed-set tex-independent speaker identification system (CISI)

Procedia PDF Downloads 309

Authors: M. Lim Haslip

Abstract:

This quantitative study explores the correlation between child gender and father-child interactions. The author analyzes data from videotaped interactions between African-American fathers and their boy or girl toddler to explain how African-American fathers and toddlers interact with each other and whether these interactions differ by child gender. The purpose of this study is to investigate the research question: 'How, if at all, do fathers’ speech and gestures differ when interacting with their two-year-old sons versus daughters during free play?' The objectives of this study are to describe how child gender impacts African-American fathers’ verbal communication, examine how fathers gesture and speak to their toddler by gender, and to guide interventions for low-income African-American families and their children in early language development. This study involves a sample of 41 low-income African-American fathers and their 24-month-old toddlers. The videotape data will be used to observe 10-minute father-child interactions during free play. This study uses the already transcribed and coded data provided by Dr. Meredith Rowe, who did her study on the impact of African-American fathers’ verbal input on their children’s language development. The Child Language Data Exchange System (CHILDES program), created to study conversational interactions, was used for transcription and coding of the videotape data. The findings focus on the quantity of speech, diversity of speech, complexity of speech, and the quantity of gesture to inform the vocabulary usage, number of spoken words, length of speech, and the number of object pointings observed during father-toddler interactions in a free play setting. This study will help intervention and prevention scientists understand early language development in the African-American population. It will contribute to knowledge of the role of African-American fathers’ interactions on their children’s language development. It will guide interventions for the early language development of African-American children.

Keywords: parental engagement, early language development, African-American families, quantity of speech, diversity of speech, complexity of speech and the quantity of gesture

Procedia PDF Downloads 105

Authors: Anja Kurz, Marc Flynn, Tobias Good, Marco Caversaccio, Martin Kompis

Abstract:

Bone Anchored Hearing Implants (BAHI) are routinely used in patients with conductive or mixed hearing loss, e.g. if conventional air conduction hearing aids cannot be used. New sound processors and new fitting software now allow the adjustment of parameters such as loudness compression ratios or maximum power output separately. Today it is unclear, how the choice of these parameters influences aided speech understanding in BAHI users. In this prospective experimental study, the effect of varying the compression ratio and lowering the maximum power output in a BAHI were investigated. Twelve experienced adult subjects with a mixed hearing loss participated in this study. Four different compression ratios (1.0; 1.3; 1.6; 2.0) were tested along with two different maximum power output settings, resulting in a total of eight different programs. Each participant tested each program during two weeks. A blinded Latin square design was used to minimize bias. For each of the eight programs, speech understanding in quiet and in noise was assessed. For speech in quiet, the Freiburg number test and the Freiburg monosyllabic word test at 50, 65, and 80 dB SPL were used. For speech in noise, the Oldenburg sentence test was administered. Speech understanding in quiet and in noise was improved significantly in the aided condition in any program, when compared to the unaided condition. However, no significant differences were found between any of the eight programs. In contrast, on a subjective level there was a significant preference for medium compression ratios of 1.3 to 1.6 and higher maximum power output.

Keywords: Bone Anchored Hearing Implant, baha, compression, maximum power output, speech understanding

Procedia PDF Downloads 387

Authors: Wale Agbaje

Abstract:

The aim of this research study is to assess the impact of financial statement fraud on profitability of some selected Nigerian manufacturing firms covering (2002-2016). The specific objectives focused on to ascertain the effect of incorrect asset valuation on return on assets (ROA) and to ascertain the relationship between improper expense recognition and return on assets (ROA). To achieve these objectives, descriptive research design was used for the study while secondary data were collected from the financial reports of the selected firms and website of security and exchange commission. The analysis of covariance (ANCOVA) was used and STATA II econometric method was used in the analysis of the data. Altman model and operating expenses ratio was adopted in the analysis of the financial reports to create a dummy variable for the selected firms from 2002-2016 and validation of the parameters were ascertained using various statistical techniques such as t-test, co-efficient of determination (R2), F-statistics and Wald chi-square. Two hypotheses were formulated and tested using the t-statistics at 5% level of significance. The findings of the analysis revealed that there is a significant relationship between financial statement fraud and profitability in Nigerian manufacturing industry. It was revealed that incorrect assets valuation has a significant positive relationship and so also is the improper expense recognition on return on assets (ROA) which serves as a proxy for profitability. The implication of this is that distortion of asset valuation and expense recognition leads to decreasing profit in the long run in the manufacturing industry. The study therefore recommended that pragmatic policy options need to be taken in the manufacturing industry to effectively manage incorrect asset valuation and improper expense recognition in order to enhance manufacturing industry performance in the country and also stemming of financial statement fraud should be adequately inculcated into the internal control system of manufacturing firms for the effective running of the manufacturing industry in Nigeria.

Keywords: Althman's Model, improper expense recognition, incorrect asset valuation, return on assets

Procedia PDF Downloads 161

Authors: Hailang Yang

Abstract:

To ensure the confidentiality and integrity of speech signals in communication transmission, this paper proposes a voice signal encryption scheme based on chaotic theory. Firstly, the scheme utilizes chaotic mapping to generate a key stream and then employs the key stream to perform bitwise exclusive OR (XOR) operations for encrypting the speech signal. Additionally, the scheme utilizes a chaotic hash function to generate a Message Authentication Code (MAC), which is appended to the encrypted data to verify the integrity of the data. Subsequently, we analyze the security performance and encryption efficiency of the scheme, comparing and optimizing it against existing solutions. Finally, experimental results demonstrate that the proposed scheme can resist common attacks, achieving high-quality encryption and speed.

Keywords: chaotic theory, XOR encryption, chaotic hash function, Message Authentication Code (MAC)

Procedia PDF Downloads 51

Authors: Jyoti Bharti, Navneet Manjhi, M. K.Gupta, Bimi Jain

Abstract:

A reliable gait features are required to extract the gait sequences from an images. In this paper suggested a simple method for gait identification which is based on moments. Moment values are extracted on different number of frames of gray scale and silhouette images of CASIA database. These moment values are considered as feature values. Fuzzy logic and nearest neighbour classifier are used for classification. Both achieved higher recognition.

Keywords: gait, fuzzy logic, nearest neighbour, recognition rate, moments

Procedia PDF Downloads 757