Search

Forced phonetic alignment (FPA) is the task of associating a given phonetic unit to a timestamp interval in the speech waveform. Phoneticians are able mark the boundaries with precision, but as the corpus grows it becomes infeasible to do it by hand. For Brazilian Portuguese (BP) in particular, only three tools appear to perform FPA: EasyAlign, Montreal Forced Aligner (MFA), and UFPAlign. Therefore, this work aims to develop resources based on Kaldi toolkit for UFPAlign, including their release alongside all scripts under open licenses; and to bring forth a comparison to the other two aforementioned aligners. Evaluation took place in terms of the phone boundary metric over a dataset of 385 hand-aligned utterances, and results show that Kaldi-based aligners perform better overall, and that UFPAlign models are more accurate than MFA’s. Furthermore, complex deep-learning-based approaches did not seem to improve performance compared to simpler models.

Phonetic analysis of speech, in general, requires the alignment of audio samples to its phonetic transcription. This task could be performed manually for a couple of files, but as the corpus grows large it becomes unfeasibly time-consuming, which emphasizes the need for computational tools that perform such speech-phonemes forced alignment automatically. Therefore, due to the scarce availability of phonetic alignment tools for Brazilian Portuguese (BP), this work describes the evolution process towards creating a free phonetic alignment tool for BP using Kaldi, a toolkit that has been the state of the art for open-source speech recognition. Five acoustic models were trained with Kaldi and tested in phonetic alignment, where the evaluation took place in terms of the phone boundary metric. The results show that its performance is similar to some Kaldi-based aligners for other languages, and superior to an outdated phonetic aligner for BP based on HTK toolkit.

Kriging is a geostatistical interpolation technique that performs the prediction of observations in unknown locations through previously collected data. The modelling of the variogram is an essential step of the kriging process because it drives the accuracy of the interpolation model. The conventional method of variogram modelling consists of using specialized knowledge and in-depth study to determine which parameters are suitable for the theoretical variogram. However, this situation is not always possible, and, in this case, it becomes interesting to use an automatic process. Thus, this work aims to propose a new methodology to automate the estimation of theoretical variogram parameters of the kriging process. The proposed methodology is based on preprocessing techniques, data clustering, genetic algorithms, and the K-Nearest Neighbor classifier (KNN). The performance of the methodology was evaluated using two databases, and it was compared to other optimization techniques widely used in the literature. The impacts of the clustering step on the stationary hypothesis were also investigated with and without trend removal techniques. The results showed that, in this automated proposal, the clustering process increases the accuracy of the kriging prediction. However, it generates groups that might not be stationary. Genetic algorithms are easily configurable with the proposed heuristic when setting the variable ranges in comparison to other optimization techniques, and the KNN method is satisfactory in solving some problems caused by the clustering task and allocating unknown points into previously determined clusters.

Computer systems are approaching human behavior in the sense of performing similar tasks, such as listening, understanding, thinking, and speaking. Although the forms of interaction with these systems have also followed the path of technology evolution, computer mice and keyboards do continue to play the leading role as a bridge between human and computer. The design of these devices forces the user to use their hands and this represents a major problem that has been already recognized in the literature with several proposed solutions. Nevertheless, in spite of the various methods found, it is difficult to find papers that make comparisons between them. Therefore, this work proposes an evaluation, using quantitative and qualitative analyses, of three alternative methods for mouse click: dwell time, mouth-puffing and electromyography. As result, both analyses showed that the interactions based on mouth-puffing and electromyography performed better than the dwell time method.

Utterance copy, also known as speech imitation, is the task of estimating the parameters of an input, target speech signal in order to artificially reconstruct another signal with the same properties at the output. This can be considered a difficult inverse problem, since the input-output relationship is often non-linear, apart from having several parameters to be estimated and adjusted. This work describes the development of an application that uses a long short-term memory neural network (LSTM) to learn how to estimate the input parameters of thel formant-based Klatt speech synthesizer. Formant-based synthesizers do not reach state-of-art performance for text-to-speech (TTS) applications, but are an important tool for linguists studies due to the high interpretability of its input parameters. The proposed system was compared to the WinSnoori baseline software on both artificially-produced target utterances, generated by the DECtalk TTS system; and natural target ones. Results show that our system outperforms the baseline for synthetic voices on the metrics of PESQ, SNR, RMSE and LSD. For natural voices, the experiments indicate the need for an architecture that does not depend on labeled data, such as reinforcement learning.

Computer mice and keyboards have for long been the dominant input devices for desktop computers. Nonetheless, they still represent quite a barrier — especially for people whose upper-limb motor skills are compromised. This work proposes a low-cost, mouth-puffing-based device that is able to perform the mouse click on computational platforms. Additionally, a driver was developed to allow the device to communicate with the computer via audio jack P2 interface. The mouse cursor is controlled by head tracking thanks to eViacam software. The evaluation took place by comparing the proposed equipment with the dwell time method on navigation tasks elaborated over websites.

Assembling metagenomic data sequenced by NGS platforms poses significant computational challenges, especially due to large volumes of data, sequencing errors, and variations in size, complexity, diversity and abundance of organisms present in a given metagenome. To overcome these problems, this work proposes an open-source, bioinfor-matic tool called GCSplit, which partitions metagenomic sequences into subsets using a computationally inexpensive metric: the GC content. Experiments performed on real data show that preprocessing short reads with GCSplit prior to assembly reduces memory consumption and generates higher quality results, such as an increase in the N50 metric and the reduction in both the L50 value and the total number of contigs produced in the assembly. GCSplit is available at https://github.com/mirand863/gcsplit.

Hereditary syndromes are responsible for 10 % of gynaecologic cancers, among which hereditary breast-ovarian cancer and hereditary non-polyposis colon cancer syndromes, known as HBOC and Lynch syndromes respectively, present the highest relative risk. The latter predisposes to endometrial cancer and both contribute to ovarian cancer. Cowden syndrome-related endometrial cancer and the increased risk of ovarian, uterine and cervical cancers associated with Peutz-Jeghers syndrome, are also demonstrated, while Li-Fraumeni syndrome patients are prone to develop ovarian and endometrial cancers. Despite these syndromes’ susceptibility to gynaecologic cancers being consensual, it is still not clear whether these tumours have any epidemiologic, clinical, pathologic or imaging specific features that could allow any of the intervening physicians to raise suspicion of a hereditary syndrome in patients without known genetic risk. Moreover, controversy exists regarding both screening and surveillance schemes. Our literature review provides an updated perspective on the evidence-based specific features of tumours related to each of these syndromes as well as on the most accepted screening and surveillance guidelines. In addition, some illustrative cases are presented. • HBOC syndrome is mainly associated with ovarian HGSC, which arises in fallopian fimbriae. • LS-related endometrial tumours show histological diversity and predilection for lower uterine segment. • LS and CS-related ovarian cancers are mostly of non-serous type, usually endometrioid. • Ovarian SCTAT and cervical adenoma malignum are strongly associated with PJS. • Unfortunately, hereditary gynaecologic cancers do not seem to have distinctive imaging features.

Technological developments converge to make people interact with electronic devices in an easy way. For people with disabilities, however, that interaction becomes something more than simple: it becomes possible. The current work presents a proposal of an open-source, low cost universal remote control system that translates user's head poses into commands to electronic devices. In addition, a proximity sensor circuit was combined to radio-frequency modules in order to act as a wireless switch. Among the applications found with respect to alternative remote controls, none of them supports head gestures as input, which would certainly make them a viable option to people whose upper limbs are compromised. A mean opinion score questionnaire was applied to volunteers in order to evaluate the system. The results show great interest of the users in this kind of technology.

Acute renal infarction occurs usually secondary to thromboembolism rather than in situ thrombosis, with atrial fibrillation being the main predisposing factor. Its non-specific presentation, being similar to renal colic and pyelonephritis, often leads to diagnostic and treatment delays. Prompt diagnosis and treatment are crucial for the outcome, as the precise therapeutic window remains unclear. Renal Doppler ultrasound is the optimal initial diagnostic study, which, if inconclusive, should be followed by contrast-enhanced CT. Despite the lack of specific guidelines, treatment is mainly based on anticoagulation and percutaneous endovascular revascularization therapy; the latter includes pharmacomechanical thrombectomy; intra-arterial thrombolysis alone or in combination with thrombus aspiration; and angioplasty, with or without stenting. We present a case of renal thromboembolism diagnosed early during the postoperative period in a patient with a single functioning kidney. Renal arterial flow restoration was achieved by means of transcatheter thrombus aspiration. This is the first report showing the effectiveness of this procedure alone as an alternative to those used so far.

An automatic speech recognition system has modules that depend on the language and, while there are many public resources for some languages (e.g., English and Japanese), the resources for Brazilian Portuguese (BP) are still limited. This work describes the development of resources and free tools for BP speech recognition, consisting of text and audio corpora, phonetic dictionary, grapheme-to-phone converter, language and acoustic models. All of them are publicly available and, together with a proposed application programming interface, have been used for the development of several new applications, including a speech module for the OpenOffice suite. Performance tests are presented, comparing the developed BP system with a commercial software. The paper also describes an application that uses synthesis and speech recognition together with a natural language processing module dedicated to statistical machine translation. This application allows the translation of spoken conversations from BP to English and vice versa. The resources make easier the adoption of BP speech technologies by other academic groups and industry.

Mobile Augmented Reality has become more popular mainly because computational resources available in mobile devices, and in the enhanced view of real world that can be seen by the user. The interaction becomes an important point for success of these applications, featuring a natural and intuitive way for the user, and the chance of one, or two hands free for other interaction or activity. Therefore, this work presents the usability analysis of a Mobile Augmented Reality, the ARGuide, with the use of a voice service for interaction. The usability test analysis is based upon the most common interaction tasks of the users in this type of application. In the end, some good practices for the application interface building and speech interaction are shown.

A linfohistiocitose hemofagocitica caracteriza-se por ativacao e proliferacao excessiva de linfocitos e macrofagos. Quando associada a artrite idiopatica juvenil e tambem conhecida por "sindrome de ativacao do macrofago", sendo uma complicacao potencialmente fatal desta doenca. Apresentamos o caso de uma mulher de 26 anos portadora de artrite idiopatica juvenil (poliartrite, fator reumatoide negativo), com diagnostico aos 13 anos, em uso de antiinflamatorios nao esteroidais (diclofenaco, nimesulide). Admitida com quadro de resposta inflamatoria sistemica, febre, linfonodomegalia, esplenomegalia, anemia, trombocitopenia, hipofibrinogenemia, hiperferritinemia, hipertrigliceridemia e achados de hematofagocitose na medula ossea. Os autores discutem aspectos relacionados com a patogenese, diagnostico e tratamento desta doenca pouco conhecida.

Active learning is a type of semi-supervised learning in which the training algorithm is able to obtain the labels of a small portion of the unlabeled dataset by interacting with an external source (e.g. a human annotator). One strategy employed in active learning is based on the exploration of the cluster structure in the data, by using the labels of a few representative samples in the classification of the remaining points. In this paper we show that unsupervised feature learning can improve the "purity" of clusters found, and how this can be combined with a simple but effective active learning strategy. The proposed method shows state-of-the art performance in MNIST digit recognition in the semi-supervised setting.

Although most with Chagas Disease (CD) stay asymptomatic indefinitely, about one third of them progress to chronic cardiomyopathy. Since mannose-binding lectin (MBL) seems to have a relevant role in tissue injury present in inflammatory cardiac disorders we aimed to investigate a possible association of MBL levels with the susceptibility and clinical progression of CD. A total of 150 CD patients presenting indeterminate, cardiac, digestive and associated forms and 139 controls were studied. MBL levels were determined by ELISA and C-reactive protein (hs-CRP) by nephelometry. High MBL values (>1000ng/ml) were associated with the severity of CD cardiomyopathy in both patients with echocardiographic alterations [O.R=25.88; I.C. 95%: 3.07–217.9; p =0.00013] and cardiac insufficiency [O.R=2.91; I.C. 95%: 1.08–7.80; p =0.02] when compared to patients without those conditions. Increased MBL levels may contribute to the inflammatory process and myocardial injury in chronic CD.

Soybean [(Glycine max (L.) Merrill] commercial fields, maintained under a no-till system, were sampled during the crop seasons 1990/91 and 1996/97, in Cruz Alta and Júlio de Castilhos, Rio Grande do Sul State, to determine the distribution of eggs, larvae and adults of Sternechus subsignatus Boheman (Coleoptera : Curculionidae). Soybean plants and soil were examined at different time schedules. Eggs and larvae were recorded in main stems, lateral branches and leaf petioles, divided into the upper third, medium and lower third of soybean plants. Eggs (87%), and larvae (79%) were mainly observed in the main stem and in the medium plant sections (87% and 78%, respectively). Larval movement was not observed because larvae remained inside the galls, in the same area where the eggs were laid. Adults were located in different places depending on the sampling time. At night, adults were observed to move to the upper third of the plant and, during the day, down to the lower third of the plant and to the soil. The insects are normally concealed during their pre- and post- embryonic stage throughout most of the day. Estudou-se a localização e a distribuição de ovos, larvas e adultos de Sternechus subsignatus Boheman (Coleoptera: Curculionidae) em plantas de soja [(Glycine max (L.) Merrill] e no solo, em diferentes horários do dia, em lavouras comerciais no sistema plantio direto, entre as safras agrícolas de 1990/91 a 1996/97, em Cruz Alta e Júlio de Castilhos (RS). Ovos e larvas foram encontrados na haste principal, nos ramos laterais e nos pecíolos, nas partes superior, mediana e inferior das plantas. Não se verificou movimento das larvas para as partes mais altas ou mais baixas das planta à medida que se desenvolveram. Ovos (90%) e a maioria das larvas (87%) foram encontrados na haste principal e na parte mediana das plantas (79% e 78%, respectivamente). Adultos foram encontrados no terço superior, médio e inferior das plantas e no solo, em todos os horários de coleta. Verificou-se movimentação dos adultos para as partes mais altas das plantas durante a noite, e para as partes mais baixas das plantas e para o solo, durante o dia. Esses resultados evidenciam que a maioria dos ovos e das larvas fica protegida no interior da haste principal e na parte mediana das plantas, sob a folhagem. Adultos, durante o dia, ficam escondidos nas partes mais baixas das plantas, entre as folhas e no solo.

Dietary supplements containing ephedrine and other alkaloids related to ephedrine are largely consumed in various countries, with the purpose of energetic stimulation and weight loss. Despite the fact that it is not approved for marketing in Brazil, these products may be freely purchased over the Internet or at gyms/fitness centers. We report the case of a young athlete with no risk factors for cardiovascular disease who experienced a myocardial infarction during the period in which he used an ephedrine-rich supplement.

Ganglion and synovial cysts occur mainly, but not necessarily, in association with osteoarthritis. Presentation varies widely, ranging from small, incidentally detected, asymptomatic lesions to giant ones that might be the source of symptoms, either due to their compressive effect on adjacent structures or due to complications, such as rupture. On magnetic resonance imaging they are typically presented as smooth, well-circumscribed, thin-walled, unilocular, and homogeneously T2-hyperintense lesions. An identifiable thin stalk communicating to the joint space is not infrequent. Nevertheless, depending on their age, anatomic location, and eventual complication, they might have many distinct appearances, including septae and internal debris, which the radiologist must be familiar with in order to accurately differentiate them from worrisome cystic-like lesions. With regard to this diversity, some illustrative cases are presented.

In recent years, the performance of personal computers has evolved with the production of ever faster processors, a fact that enables the adoption of speech processing in computer-assisted education. There are several speech technologies that are effective in education, among which text-to-speech (TTS) and automatic speech recognition (ASR) are the most prominent. TTS systems [45] are software modules that convert natural language text into synthesized speech. ASR [18] can be seen as the TTS inverse process, in which the digitized speech signal, captured for example via a microphone, is converted into text.