搜索结果：Zero-copy

Autonomous vehicles, open-world robots, and other automated systems rely on accurate, efficient perception modules for real-time object detection. Although high-precision models improve reliability, their processing time and computational overhead can hinder real-time performance and raise safety concerns. This paper introduces an Edge-based Mixture-of-Experts Optimal Sensing (EMOS) System that addresses the challenge of co-achieving accuracy, latency and scene adaptivity, further demonstrated in the open-world autonomous driving scenarios. Algorithmically, EMOS fuses multimodal sensor streams via an Adaptive Multimodal Data Bridge and uses a scenario-aware MoE switch to activate only a complementary set of specialized experts as needed. The proposed hierarchical backpropagation and a multiscale pooling layer let model capacity scale with real-world demand complexity. System-wise, an edge-optimized runtime with accelerator-aware scheduling (e.g., ONNX/TensorRT), zero-copy buffering, and overlapped I/O-compute enforces explicit latency/accuracy budgets across diverse driving conditions. Experimental results establish EMOS as the new state of the art: on KITTI, it increases average AP by 3.17% while running $2.6\times$2.6× faster on Nvidia Jetson. On nuScenes, it improves accuracy by 0.2% mAP and 0.5% NDS, with 34% fewer parameters and a $15.35\times$15.35× Nvidia Jetson speedup. Leveraging multimodal data and intelligent experts cooperation, EMOS delivers accurate, efficient and edge-adaptive perception system for autonomous vehicles, thereby ensuring robust, timely responses in real-world scenarios.

In light of the increasing threat posed by cyberattacks, it is imperative for organizations to accurately identify malicious network traffic. However, the imbalance among various attack categories diminishes the accuracy of model predictions. To address this issue, we propose the Maple-IDS dataset as an innovative solution. We utilize DPDK along with its zero-copy (ZC) technology and BPF compiler to compile filtering rules. Additionally, a headless client is employed to generate control traffic, thereby preventing overfitting. Our data collections are sourced from a variety of operating systems and middleware platforms, ensuring broad applicability and relevance. By comparing our dataset with the CIC-IDS-2017 dataset, we achieve a more balanced representation of attack data, which enhances the model's learning performance. To tackle the challenges of low accuracy and slow convergence speed in existing network security situation predictions, we propose a network situation awareness prediction model that integrates a residual network with an improved attention mechanism. This model leverages the attention mechanism to assign greater weight to abnormal data, thereby facilitating the accurate identification of anomalies within large data streams. Furthermore, the residual network accelerates convergence speed, enhances the model's expressive capability, and improves the efficiency of rapid response to attacks. Experimental results indicate that the accuracy of predicting attack data flows reaches an impressive 99.83%, which significantly aids in the early detection of network security threats and enables preemptive measures to maintain normal network operations.

Inference using deep neural networks on mobile devices has been an active area of research in recent years. The design of a deep learning inference framework targeted for mobile devices needs to consider various factors, such as the limited computational capacity of the devices, low power budget, varied memory access methods, and I/O bus bandwidth governed by the underlying processor's architecture. Furthermore, integrating an inference framework with time-sensitive applications - such as games and video-based software to perform tasks like ray tracing denoising and video processing - introduces the need to minimize data movement between processors and increase data locality in the target processor. In this paper, we propose Shader Neural Network (ShaderNN), an OpenGL-based, fast, and power-efficient inference framework designed for mobile devices to address these challenges. Our contributions include the following: (1) the texture-based input/output provides an efficient, zero-copy integration with real-time graphics pipelines or image processing applications, thereby saving expensive data transfers between CPU and GPU, which are unavoidable in most existing inference engines; (2) we are the first to leverage fragment shaders based on the OpenGL backend in neural network inference operators, which has an advantage in deploying parametrically small neural network models; (3) a hybrid implementation of the compute shader and fragment shader is proposed that enables layer-level shader selection to boost performance; and (4) we utilize OpenGL features - such as normalization, interpolation and texture padding - to improve performance. Experiments illustrate the favorable performance of ShaderNN over other popular on-device deep learning frameworks such as TensorFlow-Lite on the latest mobile devices powered by Qualcomm and MediaTek chips. A case study further demonstrates the usability and integration of the ShaderNN framework with a media processing Android application seamlessly. ShaderNN is available open source at Github (https://github.com/inferenceengine/shadernn).

New low-coverage single-cell DNA sequencing technologies enable the measurement of copy number profiles from thousands of individual cells within tumors. From this data, one can infer the evolutionary history of the tumor by modeling transformations of the genome via copy number aberrations. Copy number aberrations alter multiple adjacent genomic loci, violating the standard phylogenetic assumption that loci evolve independently. Thus, specialized models to infer copy number phylogenies have been introduced. A widely used model is the copy number transformation (CNT) model in which a genome is represented by an integer vector and a copy number aberration is an event that either increases or decreases the number of copies of a contiguous segment of the genome. The CNT distance between a pair of copy number profiles is the minimum number of events required to transform one profile to another. While this distance can be computed efficiently, no efficient algorithm has been developed to find the most parsimonious phylogeny under the CNT model. We introduce the zero-agnostic copy number transformation (ZCNT) model, a simplification of the CNT model that allows the amplification or deletion of regions with zero copies. We derive a closed form expression for the ZCNT distance between two copy number profiles and show that, unlike the CNT distance, the ZCNT distance forms a metric. We leverage the closed-form expression for the ZCNT distance and an alternative characterization of copy number profiles to derive polynomial time algorithms for two natural relaxations of the small parsimony problem on copy number profiles. While the alteration of zero copy number regions allowed under the ZCNT model is not biologically realistic, we show on both simulated and real datasets that the ZCNT distance is a close approximation to the CNT distance. Extending our polynomial time algorithm for the ZCNT small parsimony problem, we develop an algorithm, Lazac, for solving the large parsimony problem on copy number profiles. We demonstrate that Lazac outperforms existing methods for inferring copy number phylogenies on both simulated and real data.

A novel approach was employed to assess the contribution of the renin-angiotensin system (RAS) to obstructive nephropathy in neonatal mice having zero to four functional copies of the angiotensinogen gene (Agt). Two-day-old mice underwent unilateral ureteral obstruction (UUO) or sham operation; 28 days later, renal interstitial fibrosis and tubular atrophy were quantitated. In all Agt genotypes, UUO reduced ipsilateral renal mass and increased that of the opposite kidney. Renal interstitial collagen increased after UUO linearly with Agt expression, from a fractional area of 25% in zero-copy mice to 54% in two-copy mice. Renal expression of transforming growth factor-beta1 was increased by ipsilateral UUO in mice expressing Agt, but not in zero-copy mice. However, the prevalence of atrophic tubules due to UUO did not vary with Agt expression. Blood pressure was not different in all groups, except for a reduction in sham zero-copy mice. We conclude that a functional RAS is not necessary for compensatory renal growth. This study demonstrates conclusively that angiotensin regulates at least 50% of the renal interstitial fibrotic response in obstructive nephropathy, an effect independent of systemic hemodynamic changes. Angiotensin-induced fibrosis likely is a mechanism common to the progression of many forms of renal disease.

We have analyzed the survival motor neuron gene (SMN1) dosage in 100 parents of children with homozygous SMN1 deletions. Of these parents, 96 (96%) demonstrated the expected one-copy SMN1 carrier genotype. However, four parents (4%) were observed to have a normal two-copy SMN1 dosage. The presence of two intact SMN1 genes in the parent of an affected child indicates either the occurrence of a de novo mutation event or a situation in which one chromosome has two copies of SMN1, whereas the other is null. We have separated individual chromosomes from two of these parents with two-copy SMN1 dosage by somatic cell hybridization and have employed a modified quantitative dosage assay to provide direct evidence that one parent is a two-copy/ zero-copy SMN1 carrier, whereas the other parent had an affected child as the result of a de novo mutation. These findings are important for assessing the recurrence risk of parents of children with spinal muscular atrophy and for providing accurate family counseling.

Essential hypertension probably results from combinations of small genetic variations that are partly normal variations and may not be appreciably harmful individually. Strategies to identify genes contributing to hypertension are discussed in this review. Gene targeting approaches, especially gene titration, have been used in these studies of hypertension. Gene titration experiments vary the expression of a chosen gene product by generating animals having different numbers of copies of the gene coding for the product. Gene titration is powerful for analyzing quantitative variations seen in common polygenic disorders, such as kidney diseases, diabetes mellitus, and atherosclerosis, as well as hypertension, because it allows tests of causation by determining the effects on a phenotype by changes in expression of the altered gene and because it matches normal quantitative variations more closely than is possible with classic transgenic mice. The use of zero-copy (gene "knockout") animals generated by gene disruption for studies of qualitative gene effects is also discussed. These various gene targeting experiments help identify genes regulating BP, promote a better understanding of the pathophysiology of the condition, and help identify potential targets for therapies.

By using standard restriction fragment length polymorphism, 6 zero-copy IS6110 Mycobacterium tuberculosis isolates were identified from 1180 Maryland isolates as part of the National Tuberculosis Genotyping and Surveillance Network Project. By using various genotyping methods, we demonstrated that this zero band cluster can be differentiated into six genotypes.

Recombinant bacterial cells in a fermentation broth rarely contain the same number of plasmids, even though this simplification is often used. Recent work has however indicated limitations of the simplified approach. Based on these studies, the distribution of plasmid copy numbers per cell has been represented macroscopically here in a Gaussian form for the fraction of biomass as a function of the copy number. Applying this distribution and an experimentally validated kinetic model to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) synthesis by Escherichia coli containing the plasmid pBR Eco gap, it is seen that GAPDH production in a batch fermentation is maximized by a particular initial (non-zero) copy number variance and an optimal duration. To implement this distribution in a bioreactor, it is suggested that the profile may be discretized, inocula corresponding to the mean copy number of each fraction prepared, and then combined to obtain the seed culture.

PCR has emerged as a powerful technique for detection of various pathogens including Mycobacterium tuberculosis. In present study, eighty one samples of lymph node biopsies from clinically suspected cases of tuberculous lymphadenitis were examined for AFB, culture on Löwenstein Jensen medium and simultaneous use of two PCRs targeting IS6110 and MPB64. Positivity with M. tuberculosis culture and AFB was 13.6% and 28.4% respectively. All samples culture positive for nontuberculous mycobacteria were negative by both PCR systems. Higher proportion of positive results were observed with PCR targeting IS6110 by which 56 of 81 (69.1%) samples showed positive results as compared to PCR targeting MPB64 by which 39 of 81 (48.2 %) samples showed positive results. When combined, 63 out of 81 (77.8%) samples were detected positive for M. tuberculosis DNA. However, 7/81 (8.6 %) samples remained negative by IS6110 but positive by MPB64 method. Thus our data suggest that the use of one additional PCR (other than IS6110 system) can reduce false negativity of PCR results in the samples harboring zero copy of IS6110 element which is known to exist in Indian population.

Mice lacking the gene (Npr1) encoding the natriuretic peptide receptor A (NPRA) have hypertension with elevated blood pressure and cardiac hypertrophy. In particular, Npr1 gene-deficient male mice exhibit lethal vascular events similar to those seen in untreated human hypertensive patients. Serum testosterone levels tend to be lower in hypertensive male humans than in normal males without hypertension, but the genetic basis for this tendency remains unknown. To determine whether Npr1 gene function affects the testosterone level, we measured serum testosterone in male hypertensive mice lacking a functional Npr1 gene, wild-type animals with two copies, and the gene-duplicated littermates expressing four copies of the gene. In the Npr1 gene-knockout (zero-copy) mice, the serum testosterone level was 62% lower than that in the two-copy control mice (80+/-10 ts. 120+/-14 ng/ml, respectively; P < 0.005). Serum testosterone in the four-copy mice was 144% (P < 0.005) of that in the two-copy wild-type control mice. To investigate the role of NPRA in testicular steroidogenesis, we analyzed atrial natriuretic peptide (ANP)-dependent guanylyl cyclase activation, accumulation of intracellular cGMP, and testosterone production in purified primary Leydig cells from animals with zero, two, or four copies of the Npr1 gene. Leydig cells lacking the Npr1 gene did not show ANP-stimulated guanylyl cyclase activation or cGMP accumulation and had no ANP-dependent testosterone production. ANP stimulation of Leydig cells from the four-copy males elicited a 2-fold greater production of cGMP compared to that in the two-copy wild-type counterparts (260+/-12 vs. 126+/-7 pmol/l x 10(6) cells; P < 0.001). Similarly, ANP-dependent testosterone production in Leydig cells was nearly twice as high in four-copy mice as in two-copy wild-type controls (561+/-18 vs. 325+/-11 ng/l x 10(6) cells; P < 0.001). ANP-dependent guanylyl cyclase activation and production of cGMP in Leydig cells increased progressively with the number of Npr1 gene copies. Our results establish the existence of an alternate mechanism for testicular steroidogenesis that is stimulated by NPRA-dependent cGMP signaling, in addition to that mediated by gonadotropins, via a cAMP pathway. These findings demonstrate the role of Npr1 gene function in the maintenance of serum testosterone levels and testicular steroidogenesis and provide a genetic link between hypertension associated with decreased NPRA and low testosterone levels.

In the present study, genetic diversity analysis of Mycobacterium tuberculosis isolated from patients attending a tertiary care hospital, North India, has been attempted. Eighty three isolates of M. tuberculosis were subjected to DNA fingerprinting using spoligotyping and IS6110-RFLP techniques. Spoligotype patterns showed that central Asian (32.5%), ill defined T (13.2%) and Beijing (10.8%) families were predominant in ongoing transmission of the bacterium. Two STs; ST26 (CAS_Delhi) and ST1 (Beijing) represented 36.1% of the total M. tuberculosis population in eastern Uttar Pradesh, North India. IS6110 RFLP analysis showed that isolates having low and zero copy number of the IS element were 15.6% and 19.2%, respectively. Out of the 47 isolates clustered by spoligotyping, 40 could be further differentiated as unique strains by IS6110-RFLP. Therefore, this study recommends that both the techniques be used simultaneously for DNA fingerprinting of M. tuberculosis in India.

Circulating plasma Epstein Barr Virus DNA (EBV-DNA) is a sensitive and specific marker of nasopharyngeal carcinoma (NPC). The mainstay of treatment of metastatic NPC is systemic chemotherapy and resection for solitary metastasis. Despite high response rate to chemotherapy, complete remission is uncommonly seen. We report a case of recurrent metastatic NPC in a 43-year-old man, who achieved complete remission three times with chemotherapy and surgery. Serial plasma EBV-DNA levels were measured during the course of disease. The patient had three episodes of recurrences of NPC manifested as distant metastasis. Both time, rise in the plasma EBV-DNA level preceded detection of recurrences by imaging. Following systemic chemotherapy, he achieved complete remission each time, of which was confirmed by 18-flourodeoxyglucose positron emission tomography and hepatectomy pathology. The plasma EBV-DNA level dropped to zero copy/ml at the time of each remission. This case highlights the high chemosensitivity of NPC by illustrating a rare occurrence of complete response of metastatic NPC to chemotherapy. This case also underscores the usefulness of EBV-DNA as a useful tool in monitoring NPC by its ability to detect early recurrence and excellent correlation with treatment response.

Although several studies have evaluated the association between interleukin-1B (IL1B) polymorphisms and the risk of chronic obstructive pulmonary disease (COPD), most of these studies have focused on -511C-->T and -31T-->C polymorphisms, and the results of these studies have been inconsistent. This study was conducted to investigate the association between four potentially functional polymorphisms of the IL1B gene (-3737C-->T, -1464G-->C, -511C-->T, and -31T-->C) and the risk of COPD. In addition, we examined a potential interaction of the IL1B polymorphisms with the VNTR polymorphism of the IL-1 receptor antagonist (IL1RN) gene in determining the risk of COPD. The IL1B and IL1RN genotypes were determined in 311 COPD patients and 386 healthy controls. Individuals with at least one variant allele of the -511C-->T and -31T-->C polymorphisms were at a significantly increased risk for COPD when compared to carriers with each homozygous wild-type allele [adjusted odds ratio (OR) 1.53, 95% confidence interval (CI) 1.03-2.26, P=0.03; and adjusted OR 1.50, 95% CI 1.02-2.24, P=0.04, respectively]. When the COPD cases were stratified according to disease severity, the presence of at least one -511T and -31C alleles was significantly associated with severe COPD (adjusted OR 2.80, 95% CI 1.47-5.33, P=0.002; and adjusted OR 2.33, 95% CI 1.24-4.40, P=0.01, respectively), however, there was no significant association between the -511C-->T and -31T-->C genotypes and mild-to-moderate COPD. In addition, individuals carrying at least one IL1RN*2 allele were at a significantly lower risk for COPD compared to subjects carrying no IL1RN*2 allele (adjusted OR 0.51, 95% CI 0.26-0.98, P=0.04). In haplotype/diplotype analyses, individuals with one or two copies of the IL1B CCTC haplotype that carried the risk allele at all of the -3737C-->T, -1464G-->C, -511C-->T, and -31T-->C loci, were at a significantly increased risk of severe COPD when compared with subjects with zero copy of the CCTC haplotype (adjusted OR 1.96, 95% CI 1.16-3.29, P=0.01). These findings suggest that polymorphisms in the IL1B and IL1RN genes might be useful markers for determining genetic susceptibility to COPD in a Korean population.

The association of apolipoprotein E (APOE) genotypes with bone mineral density (BMD) and risk of osteoporosis have remained unclear. The influence of APOE gene polymorphisms on BMD as genetic mediators of osteoporosis risk needs to be explored in Indian postmenopausal females where this disease is rising rampantly. The present study investigated the role and relevance of four pertinent APOE single nucleotide polymorphisms: 5'UTR G/C (rs440446), Int2 G/A (rs769450), Exon4 T/C (rs429358), Exon4C/T (rs7412) in DEXA verified 133 osteoporotic, 57 osteopenic and 83 normal postmenopausal females of India, who were not taking hormone replacement therapy. Minor allele frequencies of rs440446 and rs429358 were higher in osteoporotic females (0.31, 0.18) than osteopenic (0.29, 0.15) and females having normal bone mass (0.16, 0.07). Disease association analysis revealed a susceptibility haplotype CGTC (in order of rs440446, rs769450, rs429358, rs7412) and the carriers of this haplotype has higher risk of osteopenia (OR 3.53, 95% CI 1.21-11.0, P=0.017) and osteoporosis (OR 3.61, 95% CI 1.53-9.48, P=0.002) after adjusting the confounding effect of age, BMI and years since menopause. Females who possess either one copy or two copies of the haplotype have lesser BMD values of lumbar spine (0.88 and 0.85 g/cm(2)) and femoral neck (0.84 and 0.82 g/cm(2)) than those females who possess zero copy (0.9 and 0.87 g/cm(2), respectively). The present study exposed a susceptibility haplotype CGTC, within APOE gene, which was found to be associated with BMD and risk of osteopenia and osteoporosis in postmenopausal females of India.

CYP2D6 is an important drug-metabolizing enzyme involved in the metabolism of 20-25% of commonly prescribed drugs. Genetic polymorphism of CYP has clinically significant modifications in patients' drug-metabolizing capacities. Since gene copy number variation (CNV) and single nucleotide polymorphism (SNP) frequently occur in the CYP2D6 gene, which the activity of CYP2D6 particularly depend on the genetic factors. This study aimed to investigate the frequencies of CYP2D6 genotypes in a Japanese female subject of 216 healthy volunteers. The volunteers were genotyped for CNV Exon 9 and four CYP2D6 genetic variants (*2, *5, *10, *14, *41) performed by TaqMan&#xae; genotyping assays. The CNV allele frequencies were 82.9% for two copies, 11.6% for one copy, 4.6% for three copies and 0.9% for zero copy, respectively. The frequencies of CYP2D6*1, *2, *5, *10, *14, and *41 were 38.7, 16.7, 6.3, 34.7, 0.2, and 1.2%, respectively. CYP2D6*5 and *14 were the major defective alleles. However, this genotyping is labor intensive, time consuming, and costly. We report an optimized novel protocol for the determination of CNV and SNP in CYP2D6 gene by real-time quantitative PCR. This can lower the cost and accurately determine CNV and SNP in the CYP2D6 gene with a higher output and enabling reliable estimates of disease prediction in large epidemiological samples.

The purpose of this study was to address the association between the ataxia telangiectasia mutated (ATM) gene polymorphisms and susceptibility to DNA repair capacity (DRC) among polycyclic aromatic hydrocarbons (PAHs)-exposed workers. Polymorphisms of ATM were genotyped. DRC was determined by comet assay. Chromosomal damage was detected by cytokinesis-block micronucleus (CBMN) assay. Flow cytometry was used to detect the distributions of cell cycle. Expressions of ATM and rH2AX were determined by immunoblotting analysis. Luciferase assays were performed to determine the functional difference of ATM promoter region allele. Subjects carrying T allele of rs228589 had significantly lower DRC compared with those with AA genotype. Subjects carrying G allele of rs652311 had significantly lower DRC than those with zero copy number of haplotype CGGT. SH ataxia telangiectasia mutated (SHATM) cells had significantly lower DRC than SH green fluorescent protein (SHGFP) cells induced by bleomycin and higher CBMN frequencies treated by benzo(a)pyrene [B(a)P] than SHGFP cells. After B(a)P treatment, a decrease in the percentage of G1 phase cells was observed in SHATM cells compared with SHGFP cells, rH2AX expressions were increased in SHATM cells and SHGFP cells, but ATM expressions had no change in 16HBE-SHGFP cells and HEK-SHGFP cells. Luciferase expression was not different between rs228589T and rs228589A plasmid constructs. In conclusions, it is suggested that ATM polymorphisms are associated with DRC among PAHs-exposed workers and ATM plays key roles in repair of chromosomal damage and cell cycle control with the treatment of B(a)P.

To identify the IS6110-restriction fragment length polymorphism (RFLP) DNA fingerprinting patterns of some clinical isolates of Mycobacterium tuberculosis (MTB) isolated from Ningxia, Beijing and Shanghai in recent years, and to observe their epidemiological characteristics. Chromosomal DNA of MTB was digested with endonuclease PvuII, then electrophoresed on agarose gel plate and transferred capillarilly to a nylon filter and hybridized with 245 bp fragment of IS6110 which labeled using random primer fluorescein labeling kit. The RFLP patterns of Southern hybridization were inspected autofluorographically and the chromosomal DNAs of MTB were thereby typed. Most of 103 isolates of MTB shared 8 approximately 21 copies. Some of them were clustered. Strains isolated from Ningxia and Beijing had similar DNA fingerprinting patterns. One zero copy strain and one single copy strain were found among isolates from Shanghai. IS6110-RFLP based typing is feasible for MTB molecular epidemiological study in China. Most of isolates of MTB show analyzable patterns. Isolates of MTB from Ningxia and Beijing have close genomic relation.