Search

Since the invention of the microphone by Barina in 1876, there have been numerous applications of audio processing, such as phonographs, broadcasting stations, and public address systems, which merely capture and amplify sound and play it back. Nowadays, audio processing involves analysis and noise-filtering techniques. There are various methods for noise filtering, each employing unique algorithms, but they all require two or more microphones for signal processing and analysis. For instance, on mobile phones, two microphones located in different positions are utilized for active noise cancellation (one for primary audio capture and the other for capturing ambient noise). However, a drawback is that when the sound source is distant, it may lead to poor audio capture. To capture sound from distant sources, alternative methods, like blind signal separation and beamforming, are necessary. This paper proposes employing a beamforming algorithm with two microphones to enhance speech and implementing this algorithm on an embedded system. However, prior to beamforming, it is imperative to accurately detect the direction of the sound source to process and analyze the audio from that direction. [ABSTRACT FROM AUTHOR]

In recent years, embedded system technologies and products for sensor networks and wearable devices used for monitoring people's activities and health have become the focus of the global IT industry. In order to enhance the speech recognition capabilities of wearable devices, this article discusses the implementation of audio positioning and enhancement in embedded systems using embedded algorithms for direction detection and mixed source separation. The two algorithms are implemented using different embedded systems: direction detection developed using TI TMS320C6713 DSK and mixed source separation developed using Raspberry Pi 2. For mixed source separation, in the first experiment, the average signal-to-interference ratio (SIR) at 1 m and 2 m distances was 16.72 and 15.76, respectively. In the second experiment, when evaluated using speech recognition, the algorithm improved speech recognition accuracy to 95%. [ABSTRACT FROM AUTHOR]

With the rapid advancement of modern hardware technology, breakthroughs have been made in many areas of artificial intelligence research, leading to the direction of machine replacement or assistance in various fields. However, most artificial intelligence or deep learning techniques require large amounts of training data and are typically applicable to a single task objective. Acquiring such large training datasets can be particularly challenging, especially in domains like medical imaging. In the field of image processing, few-shot image segmentation is an area of active research. Recent studies have employed deep learning and meta-learning approaches to enable models to segment objects in images with only a small amount of training data, allowing them to quickly adapt to new task objectives. This paper proposes a network architecture for meta-learning few-shot image segmentation, utilizing a meta-learning classification weight transfer network to generate masks for few-shot image segmentation. The architecture leverages pre-trained classification weight transfers to generate informative prior masks and employs pre-trained feature extraction architecture for feature extraction of query and support images. Furthermore, it utilizes a Feature Enrichment Module to adaptively propagate information from finer features to coarser features in a top-down manner for query image feature extraction. Finally, a classification module is employed for query image segmentation prediction. Experimental results demonstrate that compared to the baseline using the mean Intersection over Union (mIOU) as the evaluation metric, the accuracy increases by 1.7% in the one-shot experiment and by 2.6% in the five-shot experiment. Thus, compared to the baseline, the proposed architecture with meta-learning classification weight transfer network for mask generation exhibits superior performance in few-shot image segmentation. [ABSTRACT FROM AUTHOR]

In the research into semantic sentiment analysis, researchers commonly use some factor rules, such as the utilization of emotional keywords and the manual definition of emotional rules, to increase accuracy. However, this approach often requires extensive data and time-consuming training, and there is a need to make the system simpler and more efficient. Recognizing these challenges, our paper introduces a new semantic sentiment analysis system designed to be both higher in quality and more efficient. The structure of our proposed system is organized into several key phases. Initially, we focus on data training, which involves studying emotions and emotional psychology. Utilizing linguistic resources such as HowNet and the Chinese Knowledge and Information Processing (CKIP) techniques, we develop emotional rules that facilitate the generation of sparse representation characteristics. This process also includes constructing a sparse representation dictionary. We can map these back to the original vector space by resolving the sparse coefficients, representing two distinct categories. The system then calculates the error compared to the original vector, and the category with the minimum error is determined. The second phase involves inputting topics and collecting relevant comments from internet forums to gather public opinion on trending topics. The final phase is data classification, where we assess the accuracy of classified issues based on our data training results. Additionally, our experimental results will demonstrate the system's ability to identify hot topics, thus validating our semantic classification models. This comprehensive approach ensures a more streamlined and effective system for semantic sentiment analysis. [ABSTRACT FROM AUTHOR]

When recording conversations, there may be multiple people talking at once. While our human ears can filter out unwanted sounds, this can be challenging for automatic speech recognition (ASR) systems, leading to reduced accuracy. To address this issue, preprocessing mechanisms such as speech separation and targeted speaker extraction are necessary to separate each person's speech. With the development of deep learning, the quality of separated speech has improved significantly. Our objective is to focus on speaker extraction, which entails implementing a primary system for speech extraction and a secondary subsystem for delivering target information. To accomplish this, we have chosen a temporal convolutional network (TCN) architecture as the foundation of our speech extraction model. A TCN enables convolutional neural networks (CNNs) to manage time series modeling, and it can be constructed in various model lengths. Furthermore, we have integrated attention enhancement into the secondary subsystem to provide the speech extraction model with comprehensive and effective target information, which helps to improve the model's ability to estimate masks. As a result, the quality of the target speaker extraction will be greatly enhanced with a more precise mask. [ABSTRACT FROM AUTHOR]

Background: A potential role of the antimicrobial peptide LL-37, which is upregulated after infection, in the pathogenesis of Alzheimer's disease (AD) was identified. However, the clinical relevance of LL-37 in AD is not clear yet. Objective: This study aims to investigate the association of circulating LL-37 with longitudinal cognitive decline and neurodegeneration among older adults with memory complaints. Methods: This cohort study recruited 357 older adults with memory complaints. Participants were followed-up for two years and the cognitive functions were assessed using the Mini-Mental State Examination (MMSE). Serum LL-37, pTau181, and tTau levels were determined at baseline. Associations of baseline LL-37 with longitudinal cognitive decline and change of neurodegenerative biomarkers were analyzed. Results: No difference was found in the slope of longitudinal cognitive decline during follow-up between the low and high LL-37 group, adjusting for age, sex, education, body mass index, APOE ɛ4 carrier status, comorbidities, and baseline MMSE scores (difference in slope: 0.226, 95% CI: –0.169 to 0.621). Higher LL-37 levels were associated with longitudinal cognitive decline, as indicated by a decrease of MMSE scores of 3 points or above during follow-up (OR = 2.11, 95% CI: 1.32 to 3.38). The high LL-37 group had larger slopes of the increase in neurofilament light (difference in slope: 3.759, 95% CI: 2.367 to 5.152) and pTau181 (difference in slope: 0.325, 95% CI: 0.151 to 0.499) than the low LL-37 group. Conclusion: These findings support an association of the antimicrobial peptide LL-37 with AD from a clinical perspective. [ABSTRACT FROM AUTHOR]

Summary: Submergence is an abiotic stress that limits agricultural production world‐wide. Plants sense oxygen levels during submergence and postsubmergence reoxygenation and modulate their responses. Increasing evidence suggests that completely submerged plants are often exposed to low‐light stress, owing to the depth and turbidity of the surrounding water; however, how light availability affects submergence tolerance remains largely unknown.Here, we showed that Arabidopsis thaliana MYB DOMAIN PROTEIN30 (MYB30) is an important transcription factor that integrates light signaling and postsubmergence stress responses.MYB DOMAIN PROTEIN30 protein abundance decreased upon submergence and accumulated during reoxygenation. Under submergence conditions, CONSTITUTIVE PHOTOMORPHOGENIC1 (COP1), a central regulator of light signaling, caused the ubiquitination and degradation of MYB30. In response to desubmergence, however, light‐induced MYB30 interacted with MYC2, a master transcription factor involved in jasmonate signaling, and activated the expression of the VITAMIN C DEFECTIVE1 (VTC1) and GLUTATHIONE SYNTHETASE1 (GSH1) gene families to enhance antioxidant biosynthesis. Consistent with this, the myb30 knockout mutant showed increased sensitivity to submergence, which was partially rescued by overexpression of VTC1 or GSH1.Thus, our findings uncover the mechanism by which the COP1‐MYB30 module integrates light signals with cellular oxidative homeostasis to coordinate plant responses to postsubmergence stress. [ABSTRACT FROM AUTHOR]

Here the problem of designing multi-UAVs formation autonomous reconfiguration is considered. Combined with three kinds of cost functions, nonlinear dynamic equations, and four inequality constraints, one nonlinear multiobjective optimization problem is constructed. After applying weighted sum method and separating all equality or inequality constraints, the former nonlinear multiobjective optimization problem can be converted into a standard nonlinear single objective optimization problem. Then the interior point algorithm is applied to solve it. Further some improvements are proposed to avoid rank deficiency of some matrices. The equivalence property between multiobjective optimization and single objective optimization through weighted sum method is proved. Finally the efficiency of the proposed strategy can be confirmed by the simulation example results.

Background: This study aims to assess the efficacy and safety of different doses of intravenous tissue-type plasminogen activator (tPA) for acute ischemic stroke (AIS) by adopting a network meta-analysis (NMA). Methods: Studies comparing different doses of tPA in AIS were identified by retrieving electronic databases. NMAs of outcome measures included favorable functional outcome with a modified Rankin scale score (mRS) of 0 or 1 at 3 months after treatment (3M-FF), the functional independence with a mRS of 0, 1, or 2 at 3 months (3M-FI), symptomatic intracranial hemorrhage (sICH) and 3-month all-cause mortality (3M-M). Symptomatic intracranial hemorrhage (sICH) and 3-month all-cause mortality (3M-M) were assessed. Probability-based ranking and surface under cumulative ranking (SUCRA) were performed to identify the best dose of tPA. Inconsistency was evaluated by node-splitting analysis and a loop-specific approach. Publication bias was analyzed by funnel plots. Results: A total of 14 studies were included in the quantitative synthesis. The NMA results revealed no difference among low (<0.7 mg/kg), moderate (0.8 mg/kg), and standard (0.9 mg/kg) doses of tPA with regard to efficacy and safety. The SUCRAs of 3M-FF and 3M-FI showed that the standard dose ranked first, the moderate dose ranked second, and the low dose ranked third. The SUCRA of sICH showed that the standard dose ranked first (78.1%), the low dose ranked second (61.0%), and the moderate dose ranked third (11.0%). The SUCRAs of 3-month mortality showed that the standard dose ranked first (73.2%), the moderate dose ranked second (40.8%), and the low dose ranked third (36.1%). No significant inconsistency was shown by node-splitting analysis and no publication bias was shown in funnel plots. Conclusion: Lower dose tPA was comparable to the standard dose with regard to efficacy and safety. Based on the SUCRA results and American Heart Association/American Stroke Association (AHA/ASA) guidelines, the standard dose was still the optimal selection for AIS. [ABSTRACT FROM AUTHOR]