Search Results (13)

Deep learning (DL) technology has shown outstanding performance in various fields such as object recognition and classification, speech recognition, and natural language processing. However, it is well known that DL models are vulnerable to data poisoning attacks, where adversaries modify or inject data samples maliciously during the training phase, leading to degraded classification accuracy or misclassification. Since data poisoning attacks keep evolving to avoid existing defense methods, security researchers thoroughly examine data poisoning attack models and devise more reliable and effective detection methods accordingly. In particular, data poisoning attacks can be realistic in an adversarial situation where we retrain a DL model with a new dataset obtained from an external source during transfer learning. By this motivation, we propose a novel defense method that partitions and inspects the new dataset and then removes malicious sub-datasets. Specifically, our proposed method first divides a new dataset into n sub-datasets either evenly or randomly, inspects them by using the clean DL model as a poisoned dataset detector, and finally removes malicious sub-datasets classified by the detector. For partition and inspection, we design two dynamic defensive algorithms: the Sequential Partitioning and Inspection Algorithm (SPIA) and the Randomized Partitioning and Inspection Algorithm (RPIA). With this approach, a resulting cleaned dataset can be used reliably for retraining a DL model. In addition, we conducted two experiments in the Python and DL environment to show that our proposed methods effectively defend against two data poisoning attack models (concentrated poisoning attacks and random poisoning attacks) in terms of various evaluation metrics such as removed poison rate (RPR), attack success rate (ASR), and classification accuracy (ACC). Specifically, the SPIA completely removed all poisoned data under concentrated poisoning attacks in both Python and DL environments. In addition, the RPIA removed up to 91.1% and 99.1% of poisoned data under random poisoning attacks in Python and DL environments, respectively. Full article

Gait event detection is essential for controlling an orthosis and assessing the patient’s gait. In this study, patients wearing an electromechanical (EM) knee–ankle–foot orthosis (KAFO) with a single IMU embedded in the thigh were subjected to gait event detection. The algorithm detected four essential gait events (initial contact (IC), toe off (TO), opposite initial contact (OIC), and opposite toe off (OTO)) and determined important temporal gait parameters such as stance/swing time, symmetry, and single/double limb support. These gait events were evaluated through gait experiments using four force plates on healthy adults and a hemiplegic patient who wore a one-way clutch KAFO and a pneumatic cylinder KAFO. Results showed that the smallest error in gait event detection was found at IC, and the largest error rate was observed at opposite toe off (OTO) with an error rate of −2.8 ± 1.5% in the patient group. Errors in OTO detection resulted in the largest error in determining the single limb support of the patient with an error of 5.0 ± 1.5%. The present study would be beneficial for the real-time continuous monitoring of gait events and temporal gait parameters for persons with an EM KAFO. Full article

Agriculture is the second most important greenhouse gas (GHG: methane (CH₄) and nitrous oxide (N₂O) emissions)-emitting sector after the energy sector. Agriculture is also recognized as the source and sink of GHGs. The share of agriculture to the global GHG emission records has been widely investigated, but the impact on our food production systems has been overlooked for decades until the recent climate crisis. Livestock production and feed, nitrogen-rich fertilizers and livestock manure application, crop residue burning, as well as water management in flood-prone cultivation areas are components of agriculture that produce and emit most GHGs. Although agriculture produces 72–89% less GHGs than other sectors, it is believed that reducing GHG emissions in agriculture would considerably lower its share of the global GHG emission records, which may lead to enormous benefits for the environment and food production systems. However, several diverging and controversial views questioning the actual role of plants in the current global GHG budget continue to nourish the debate globally. We must acknowledge that considering the beneficial roles of major GHGs to plants at a certain level of accumulation, implementing GHG mitigation measures from agriculture is indeed a complex task. This work provides a comprehensive review of agriculture-related GHG production and emission mechanisms, as well as GHG mitigation measures regarded as potential solutions available in the literature. This review also discusses in depth the significance and the dynamics of mitigation measures regarded as game changers with a high potential to enhance, in a sustainable manner, the resilience of agricultural systems. Some of the old but essential agricultural practices and livestock feed techniques are revived and discussed. Agricultural GHG mitigation approaches discussed in this work can serve as game changers in the attempt to reduce GHG emissions and alleviate the impact of climate change through sustainable agriculture and informed decision-making. Full article

Bakanae disease caused by Fusarium fujikuroi causes crop failure and yield losses in the majority of rice-growing countries. In this study, we employed a joint strategy quantitative trait locus (QTL) mapping–Genome-Wide Association Study (GWAS) to investigate novel genetic loci associated with Bakanae disease resistance using a population of 143 BC₁F₈ RILs derived from a cross between Ilpum × Tung Tin Wan Hien1. The phenotypic data from the bioassay and the genotypic data generated using a DNA chip were utilized to perform QTL mapping and GWAS study. Our results identified a novel genetic locus qBK4^T associated with Bakanae disease resistance, which was mapped on chromosome 4 and flanked by AX-116847364 (33.12 Mbp) and AX-115752415 (33.44 Mbp) markers covering a region of 324kbp. There were 34 genes in this region including Os04g55920 (encoding a zinc-finger protein, OsJAZ1), Os04g55970 (encoding AP2-like ethylene-responsive transcription factor), etc. This study proposes qBK4^T as a novel locus for Bakanae disease resistance. The identification of qBK4^T and its flanking marker information could be useful for marker-assisted breeding and functional characterization of resistance genes against bakanae disease. Full article

Nitrogen (N) is a gas and the fifth most abundant element naturally found in the atmosphere. N’s role in agriculture and plant metabolism has been widely investigated for decades, and extensive information regarding this subject is available. However, the advent of sequencing technology and the advances in plant biotechnology, coupled with the growing interest in functional genomics-related studies and the various environmental challenges, have paved novel paths to rediscovering the fundamentals of N and its dynamics in physiological and biological processes, as well as biochemical reactions under both normal and stress conditions. This work provides a comprehensive review on multiple facets of N and N-containing compounds in plants disseminated in the literature to better appreciate N in its multiple dimensions. Here, some of the ancient but fundamental aspects of N are revived and the advances in our understanding of N in the metabolism of plants is portrayed. It is established that N is indispensable for achieving high plant productivity and fitness. However, the use of N-rich fertilizers in relatively higher amounts negatively affects the environment. Therefore, a paradigm shift is important to shape to the future use of N-rich fertilizers in crop production and their contribution to the current global greenhouse gases (GHGs) budget would help tackle current global environmental challenges toward a sustainable agriculture. Full article

In rice, high radial oxygen loss (ROL) has been associated with the reduction in the activity of methanogens, therefore reducing the formation of methane (CH₄) due to the abundance in application of nitrogen (N)-rich fertilizers. In this study, we evaluated the root growth behavior and ROL rate of a doubled haploid (DH) population (n = 117) and parental lines 93-11 (P1, indica) and Milyang352 (P2, japonica) in response to iron (II) sulfide (FeS). In addition, we performed a linkage mapping and quantitative trait locus (QTL) analysis on the same population for the target traits. The results of the phenotypic evaluation revealed that parental lines had distinctive root growth and ROL patterns, with 93-11 (indica) and Milyang352 (japonica) showing low and high ROL rates, respectively. This was also reflected in their derived population, indicating that 93.2% of the DH lines exhibited a high ROL rate and about 6.8% had a low ROL pattern. Furthermore, the QTL and linkage map analysis detected two QTLs associated with the control of ROL and root area on chromosomes 2 (qROL-2-1, 127 cM, logarithm of the odds (LOD) 3.04, phenotypic variation explained (PVE) 11.61%) and 8 (qRA-8-1, 97 cM, LOD 4.394, PVE 15.95%), respectively. The positive additive effect (2.532) of qROL-2-1 indicates that the allele from 93-11 contributed to the observed phenotypic variation for ROL. The breakthrough is that the qROL-2-1 harbors genes proposed to be involved in stress signaling, defense response mechanisms, and transcriptional regulation, among others. The qPCR results revealed that the majority of genes harbored by the qROL-2-1 recorded a higher transcript accumulation level in Milyang352 over time compared to 93-11. Another set of genes exhibited a high transcript abundance in P1 compared to P2, while a few were differentially regulated between both parents. Therefore, OsTCP7 and OsMYB21, OsARF8 genes encoding transcription factors (TFs), coupled with OsTRX, OsWBC8, and OsLRR2 are suggested to play important roles in the positive regulation of ROL in rice. However, the recorded differential expression of OsDEF7 and OsEXPA, and the decrease in OsNIP2, Oscb5, and OsPLIM2a TF expression between parental lines proposes them as being involved in the control of oxygen flux level in rice roots. Full article

In recent acoustic scene classification (ASC) models, various auxiliary methods to enhance performance have been applied, e.g., subsystem ensembles and data augmentations. Particularly, the ensembles of several submodels may be effective in the ASC models, but there is a problem with increasing the size of the model because it contains several submodels. Therefore, it is hard to be used in model-complexity-limited ASC tasks. In this paper, we would like to find the performance enhancement method while taking advantage of the model ensemble technique without increasing the model size. Our method is proposed based on a mean-teacher model, which is developed for consistency learning in semi-supervised learning. Because our problem is supervised learning, which is different from the purpose of the conventional mean-teacher model, we modify detailed strategies to maximize the consistency learning performance. To evaluate the effectiveness of our method, experiments were performed with an ASC database from the Detection and Classification of Acoustic Scenes and Events 2021 Task 1A. The small-sized ASC model with our proposed method improved the log loss performance up to 1.009 and the F${}_1$-score performance by 67.12%, whereas the vanilla ASC model showed a log loss of 1.052 and an F${}_1$-score of 65.79%. Full article

The green rice leafhopper (GRH, Nephotettix cincticeps Uhler) is one of the most important insect pests causing serious damage to rice production and yield loss in East Asia. Prior to performing RNA-Seq analysis, we conducted an electrical penetration graph (EPG) test to investigate the feeding behavior of GRH on Ilpum (recurrent parent, GRH-susceptible cultivar), a near-isogenic line (NIL carrying Grh1) compared to the Grh1 donor parent (Shingwang). Then, we conducted a transcriptome-wide analysis of GRH-responsive genes in Ilpum and NIL, which was followed by the validation of RNA-Seq data by qPCR. On the one hand, EPG results showed differential feeding behaviors of GRH between Ilpum and NIL. The phloem-like feeding pattern was detected in Ilpum, whereas the EPG test indicated a xylem-like feeding habit of GRH on NIL. In addition, we observed a high death rate of GRH on NIL (92%) compared to Ilpum (28%) 72 h post infestation, attributed to GRH failure to suck the phloem sap of NIL. On the other hand, RNA-Seq data revealed that Ilpum and NIL GRH-treated plants generated 1,766,347 and 3,676,765 counts per million mapped (CPM) reads, respectively. The alignment of reads indicated that more than 75% of reads were mapped to the reference genome, and 8859 genes and 15,815,400 transcripts were obtained. Of this number, 3424 differentially expressed genes (DEGs, 1605 upregulated in Ilpum and downregulated in NIL; 1819 genes upregulated in NIL and downregulated in Ilpum) were identified. According to the quantile normalization of the fragments per kilobase of transcript per million mapped reads (FPKM) values, followed by the Student’s t-test (p < 0.05), we identified 3283 DEGs in Ilpum (1935 upregulated and 1348 downregulated) and 2599 DEGs in NIL (1621 upregulated and 978 downregulated) with at least a log₂ (logarithm base 2) twofold change (Log₂FC ≥2) in the expression level upon GRH infestation. Upregulated genes in NIL exceeded by 13.3% those recorded in Ilpum. The majority of genes associated with the metabolism of carbohydrates, amino acids, lipids, nucleotides, the activity of coenzymes, the action of phytohormones, protein modification, homeostasis, the transport of solutes, and the uptake of nutrients, among others, were abundantly upregulated in NIL (carrying Grh1). However, a high number of upregulated genes involved in photosynthesis, cellular respiration, secondary metabolism, redox homeostasis, protein biosynthesis, protein translocation, and external stimuli response related genes were found in Ilpum. Therefore, all data suggest that Grh1-mediated resistance against GRH in rice would involve a transcriptome-wide reprogramming, resulting in the activation of bZIP, MYB, NAC, bHLH, WRKY, and GRAS transcription factors, coupled with the induction of the pathogen-pattern triggered immunity (PTI), systemic acquired resistance (SAR), symbiotic signaling pathway, and the activation of genes associated with the response mechanisms against viruses. This comprehensive transcriptome profile of GRH-responsive genes gives new insights into the molecular response mechanisms underlying GRH (insect pest)–rice (plant) interaction. Full article

Potassium chlorate (KClO₃) has been widely used to evaluate the divergence in nitrogen use efficiency (NUE) between indica and japonica rice subspecies. This study investigated the transcriptional regulation of major genes involved in the NUE in rice treated with KClO₃, which acts as an inhibitor of the reducing activity of nitrate reductase (NR) in higher plants. A set of two KClO₃ sensitive nitrate reductase (NR) and two nitrate transporter (NRT) introgression rice lines (BC2F7), carrying the indica alleles of NR or NRT, derived from a cross between Saeilmi (japonica, P1) and Milyang23 (indica, P2), were exposed to KClO₃ at the seedling stage. The phenotypic responses were recorded 7 days after treatment, and samples for gene expression, physiological, and biochemical analyses were collected at 0 h (control) and 3 h after KClO₃ application. The results revealed that Saeilmi (P1, japonica) and Milyang23 (P2, indica) showed distinctive phenotypic responses. In addition, the expression of OsNR2 was differentially regulated between the roots, stem, and leaf tissues, and between introgression lines. When expressed in the roots, OsNR2 was downregulated in all introgression lines. However, in the stem and leaves, OsNR2 was upregulated in the NR introgression lines, but downregulation in the NRT introgression lines. In the same way, the expression patterns of OsNIA1 and OsNIA2 in the roots, stem, and leaves indicated a differential transcriptional regulation by KClO₃, with OsNIA2 prevailing over OsNIA1 in the roots. Under the same conditions, the activity of NR was inhibited in the roots and differentially regulated in the stem and leaf tissues. Furthermore, the transcriptional divergence of OsAMT1.3 and OsAMT2.3, OsGLU1 and OsGLU2, between NR and NRT, coupled with the NR activity pattern in the roots, would indicate the prevalence of nitrate (NO₃¯) transport over ammonium (NH₄⁺) transport. Moreover, the induction of catalase (CAT) and polyphenol oxidase (PPO) enzyme activities in Saeilmi (P1, KClO₃ resistant), and the decrease in Milyang23 (P2, KClO₃ sensitive), coupled with the malondialdehyde (MDA) content, indicated the extent of the oxidative stress, and the induction of the adaptive response mechanism, tending to maintain a balanced reduction–oxidation state in response to KClO₃. The changes in the chloroplast pigments and proline content propose these compounds as emerging biomarkers for assessing the overall plant health status. These results suggest that the inhibitory potential of KClO₃ on the reduction activity of the nitrate reductase (NR), as well as that of the genes encoding the nitrate and ammonium transporters, and glutamate synthase are tissue-specific, which may differentially affect the transport and assimilation of nitrate or ammonium in rice. Full article

In this paper, feature extraction methods are developed based on the non-negative matrix factorization (NMF) algorithm to be applied in weakly supervised sound event detection. Recently, the development of various features and systems have been attempted to tackle the problems of acoustic scene classification and sound event detection. However, most of these systems use data-independent spectral features, e.g., Mel-spectrogram, log-Mel-spectrum, and gammatone filterbank. Some data-dependent feature extraction methods, including the NMF-based methods, recently demonstrated the potential to tackle the problems mentioned above for long-term acoustic signals. In this paper, we further develop the recently proposed NMF-based feature extraction method to enable its application in weakly supervised sound event detection. To achieve this goal, we develop a strategy for training the frequency basis matrix using a heterogeneous database consisting of strongly- and weakly-labeled data. Moreover, we develop a non-iterative version of the NMF-based feature extraction method so that the proposed feature extraction method can be applied as a part of the model structure similar to the modern “on-the-fly” transform method for the Mel-spectrogram. To detect the sound events, the temporal basis is calculated using the NMF method and then used as a feature for the mean-teacher-model-based classifier. The results are improved for the event-wise post-processing method. To evaluate the proposed system, simulations of the weakly supervised sound event detection were conducted using the Detection and Classification of Acoustic Scenes and Events 2020 Task 4 database. The results reveal that the proposed system has F1-score performance comparable with the Mel-spectrogram and gammatonegram and exhibits 3–5% better performance than the log-Mel-spectrum and constant-Q transform. Full article

Chlorate resistance analysis is an effective approach commonly used to distinguish the genetic variation between Oryza sativa L. ssp. indica and japonica, and predict the nitrogen use efficiency (NUE). This study aimed at investigating the response of a doubled haploid (DH) population derived from anther culture of 93-11 × Milyang352 exposed to 0.1% potassium chlorate (KClO₃) at the seedling stage. The results revealed that the parental rice lines 93-11 (indica) and Milyang352 (japonica) showed distinctive phenotypic responses. The parental line 93-11 scored highly sensitive (0% survival) and Milyang352 scored resistant (66.7% survival) 7 days after treatment. The DH lines reflected the differential phenotypic response observed in parental lines. Interestingly, we identified a novel quantitative trait locus (QTL) for chlorate resistance on chromosome 3 (qCHR-3, 136 cM, logarithm of the odds—LOD: 4.1) using Kompetitive Allele-Specific PCR (KASP) markers. The additive effect (−11.97) and phenotypic variation explained (PVE; 14.9%) indicated that the allele from Milyang352 explained the observed phenotypic variation. In addition, shoot growth showed a significant difference between parental lines, but not root growth. Moreover, in silico analysis identified candidate genes with diverse and interesting molecular and physiological functions. Therefore, this study suggested that the QTL qCHR-3 harbors promising candidate genes that could play a role in the regulation of nitrogen metabolism in rice. Full article

The [World Economic Forum 2016, https://www.weforum.org/events/world-economicforum- annual-meeting-2016. Accessed 5 Aug 2017] predicted that the world would face the Fourth Industrial Revolution which means innovative changes through convergence of cutting-edge information and communications technologies (ICTs) such as artificial intelligence, IoT, big data and cloud computing with conventional industries. It was forecasted that such innovation would take place across all industries and services. In particular, the Fourth Industrial Revolution has paid attention to changes in conventional traditional industries. The weak signal analysis is a method which can detect the sign of future changes at an early stage. A weak signal is defined as an advanced indicator in such changes. Therefore, its search and monitoring can be an early warning on threat and great opportunity to be prepared. To overcome the limitation of future prediction which is mostly dependent upon an expert group’s intuitive judgment, this study attempted to investigate the weak signal of convergence among industries, using knowledge structure based approaches. As new products and services can lead to convergence between industries, research data has gathered information on the products and services of global listed companies which have been established for the past 11 years. For comparative analysis, they were grouped every 5 years, and product and service trends were compared. And we detected weak signals of convergence with different industries and noticed the changes of existing industries. Historically, the convergence of industries is mainly discussed in relation to science and technology-led industries, so the industrial field that has been studied has chosen Korea’s growth engine industries, a high-tech science and technology industries. Full article

This study uses text and data mining to investigate the relationship between the text patterns of annual reports published by US listed companies and sales performance. Taking previous research a step further, although annual reports show only past and present financial information, analyzing text content can identify sentences or patterns that indicate the future business performance of a company. First, we examine the relation pattern between business risk factors and current business performance. For this purpose, we select companies belonging to two categories of US SIC (Standard Industry Classification) in the IT sector, 7370 and 7373, which include Twitter, Facebook, Google, Yahoo, etc. We manually collect sales and business risk information for a total of 54 companies that submitted an annual report (Form 10-K) for the last three years in these two categories. To establish a correlation between patterns of text and sales performance, four hypotheses were set and tested. To verify the hypotheses, statistical analysis of sales, statistical analysis of text sentences, sentiment analysis of sentences, clustering, dendrogram visualization, keyword extraction, and word-cloud visualization techniques are used. The results show that text length has some correlation with sales performance, and that patterns of frequently appearing words are correlated with the sales performance. However, a sentiment analysis indicates that the positive or negative tone of a report is not related to sales performance. Full article