Search

In speech emotion recognition (SER), our research addresses the critical challenges of capturing and evaluating node information and their complex interrelationships within speech data. We introduce Skip Graph Convolutional and Graph Attention Network (SkipGCNGAT), an innovative model that combines the strengths of skip graph convolutional networks (SkipGCNs) and graph attention networks (GATs) to address these challenges. SkipGCN incorporates skip connections, enhancing the flow of information across the network and mitigating issues such as vanishing gradients, while also facilitating deeper representation learning. Meanwhile, the GAT in the model assigns dynamic attention weights to neighboring nodes, allowing SkipGCNGAT to focus on both the most relevant local and global interactions within the speech data. This enables the model to capture subtle and complex dependencies between speech segments, thus facilitating a more accurate interpretation of emotional content. It overcomes the limitations of previous single-layer graph models, which were unable to effectively represent these intricate relationships across time and in different speech contexts. Additionally, by introducing a pre-pooling SkipGCN combination technique, we further enhance the ability of the model to integrate multi-layer information before pooling, improving its capacity to capture both spatial and temporal features in speech. Furthermore, we rigorously evaluated SkipGCNGAT on the IEMOCAP and MSP-IMPROV datasets, two benchmark datasets in SER. The results demonstrated that SkipGCNGAT consistently achieved state-of-the-art performance. These findings highlight the effectiveness of the proposed model in accurately recognizing emotions in speech, offering valuable insights and a solid foundation for future research on capturing complex relationships within speech signals for emotion recognition. [ABSTRACT FROM AUTHOR]

Coronavirus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), emerged as a global outbreak in 2019, profoundly affecting both human health and the global economy. Various vaccine modalities were developed and commercialized to overcome this challenge, including inactivated vaccines, mRNA vaccines, adenovirus vector-based vaccines, and subunit vaccines. While intramuscular vaccines induce high IgG levels, they often fail to stimulate significant mucosal immunity in the respiratory system. We employed the Newcastle disease virus (NDV) vector expressing the spike protein of the SARS-CoV-2 Beta variant (rK148/beta-S), and evaluated the efficacy of intranasal vaccination with rK148/beta-S in K18-hACE2 transgenic mice. Intranasal vaccination with a low dose (106.0 EID50) resulted in an 86% survival rate after challenge with the SARS-CoV-2 Beta variant. Administration at a high dose (107.0 EID50) led to a reduction in lung viral load and 100% survival against the SARS-CoV-2 Beta and Delta variants. A high level of the SARS-CoV-2 spike-specific IgA was also induced in vaccinated mice lungs following the SARS-CoV-2 challenge. Our findings suggest that rK148/beta-S holds promise as an intranasal vaccine candidate that effectively induces mucosal immunity against SARS-CoV-2. [ABSTRACT FROM AUTHOR]

The ongoing global spread of Newcastle disease underscores the crucial need for continued research on the efficacy of current vaccines against various circulating strains of Newcastle disease virus (NDV). The fusion gene of a representative Egyptian genotype VII.1.1 strain was used to substitute its corresponding gene in the K148/08 vaccinal strain after site directly mutating its cleavage site from 112RRQKRF117 to 112GKQGRL117. Fusion gene exchange between GVII and GI did not affect the thermostability of GI K148/08. Attenuation of the rescued virus was confirmed by mean death time 144 h with an intracerebral pathogenicity index of 0.00. Survival analysis after the challenge experiment confirmed that 107 EID50 was the protective dose of rK148/GVII-F. The haemagglutination inhibition level of antibodies required for full clinical protection was > 3.3 log2 for rK148/GVII-F and > 4.1 log2 for both K148/08 and LaSota. Oropharyngeal viral shedding was reduced on the 5th and 7th days post-challenge in the rK148/GVII-F vaccinated group. Replication and tropism investigations confirmed the respirotropic nature of LaSota, enterotropic nature of K148/08, and further attenuation of rK148/GVII-F. Altogether, rK148/GVII-F is a thermostable, safe, effective, and genetically stable vaccine candidate that could be adequate for use in countries that encounter GVII.1.1 and in those with tropical climate, such as most Middle Eastern countries. A thermostable, safe, and effective NDV GVII recombinant vaccine was generated. Fusion gene replacement with GVII did not affect GI K148/08 virus thermostability. Strain rK148/GVII-F provided adequate protection against a lethal NDV challenge. Oropharyngeal shedding was significantly reduced on post-challenge days 5 and 7. [ABSTRACT FROM AUTHOR]

Due to the immutability of blockchain, the integration with big-data systems creates limitations on redundancy, scalability, cost, and latency. Additionally, large amounts of invaluable data result in the waste of energy and storage resources. As a result, the demand for data deletion possibilities in blockchain has risen over the last decade. Although several prior studies have introduced methods to address data modification features in blockchain, most of the proposed systems need shorter deletion delays and security requirements. This study proposes a novel blockchain architecture called Unlichain that provides data-modification features within public blockchain architecture. To achieve this goal, Unlichain employed a new indexing technique that defines the deletion time for predefined lifetime data. The indexing technique also enables the deletion possibility for unknown lifetime data. Unlichain employs a new metadata verification consensus among full and meta nodes to avoid delays and extra storage usage. Moreover, Unlichain motivates network nodes to include more transactions in a new block, which motivates nodes to scan for expired data during block mining. The evaluations proved that Unlichain architecture successfully enables instant data deletion while the existing solutions suffer from block dependency issues. Additionally, storage usage is reduced by up to 10%. [ABSTRACT FROM AUTHOR]

As the utilization of complex and heavy applications increases in autonomous driving, research on using mobile edge computing and task offloading for autonomous driving is being actively conducted. Recently, researchers have been studying task offloading algorithms using artificial intelligence, such as reinforcement learning or partial offloading. However, these methods require a lot of training data and critical deadlines and are weakly adaptive to complex and dynamically changing environments. To overcome this weakness, in this paper, we propose a novel task offloading algorithm based on Lyapunov optimization to maintain the system stability and minimize task processing delay. First, a real-time monitoring system is built to utilize distributed computing resources in an autonomous driving environment efficiently. Second, the computational complexity and memory access rate are analyzed to reflect the characteristics of the deep learning applications to the task offloading algorithm. Third, Lyapunov and Lagrange optimization solves the trade-off issues between system stability and user requirements. The experimental results show that the system queue backlog remains stable, and the tasks are completed within an average of 0.4231 s, 0.7095 s, and 0.9017 s for object detection, driver profiling, and image recognition, respectively. Therefore, we ensure that the proposed task offloading algorithm enables the deep learning application to be processed within the deadline and keeps the system stable. [ABSTRACT FROM AUTHOR]

In 2020, the Y280-lineage H9N2 low-pathogenic avian influenza virus (LPAIV) was introduced into South Korea for the first time. Current vaccines are focused on the control of Y439-like viruses; however, there are continuous reports of decrease in egg production and secondary infections caused by Y280-lineage H9N2 LPAI infection in chickens. Therefore, there is an urgent need to develop effective novel vaccines against Y280-lineage H9N2 LPAI. Most commercialized avian influenza vaccines are oil-adjuvanted inactivated vaccines, which are labour-intensive to administer and require higher dosage. In this study, rK148/Y280-HA, a novel recombinant Newcastle disease virus (NDV) vectored vaccine against Y280-lineage H9N2 LPAI, was developed and evaluated using two mass-applicable administration methods, spray vaccination and drinking water vaccination. Regardless of low serum antibody haemagglutination inhibition titres against NDV and Y280-lineage H9N2 LPAI after applying the rK148/Y280-HA vaccine, vaccination with either administration method protected chickens against virulent NDV and Y280-lineage H9N2 LPAIV after the challenge. Taken together, these results indicate that the rK148/Y280 vaccine can be administered using facile mass-application methods to provide protection against the Y280-lineage LPAI. RESEARCH HIGHLIGHTS NDV vectored vaccine harbouring Y280-lineage H9N2 HA protein was successfully generated. NDV vectored vaccine provides protection against NDV. NDV vectored vaccine with H9N2 HA protects against homologous H9N2 LPAIV. [ABSTRACT FROM AUTHOR]

Discusses the collection fusion problem of image databases, concerned with retrieving relevant images by content based retrieval from image databases distributed on the Web. Focuses on a metaserver which selects image databases supporting similarity measures and proposes a new algorithm which exploits a probabilistic technique using Bayesian estimation for a linear regression model. (Author/LRW)