Search

The influence of Michigan leadership behaviors in organizational variables has been conducted in the Western context. However, limited studies have been conducted in the Vietnam context, particularly in the primary school sector. This study aims to examine the impact of Michigan leadership behaviors on organizational citizenship behaviors (OCBs) in Vietnamese primary schools. In addition, this research aims to challenge the mediating role of organizational justice (OJ) and person-organization fit (POF) in the relationship between Michigan leadership behaviors and OCBs. A total of 282 responses were collected from various primary schools in Vietnam. Exploratory factor analysis, confirmatory factor analysis, and linear regression analysis were used to answer the research question. Findings partially supported our hypotheses as two Michigan leadership dimensions significantly influence OCBs, and OJ mediates the relationship between relationship-oriented leadership behavior (ROL) and OCBs. This study opens the main gaps in the current leadership studies, including the influence of Michigan leadership behaviors on OCB, its role in promoting OJ, and the underlying mechanism through which Michigan leadership behaviors contribute to OCB with the mediation of OJ.

Abstract In the era of industry 4.0 and the widespread use of digital devices, the number of cyber attacks poses an escalating and diverse threat, jeopardizing users' online activities. Intrusion detection systems (IDS) emerge as pivotal solutions, playing a crucial role in detecting anomalous signals within network systems. To counter novel attack patterns, IDS systems require periodic rule updates for effective identification of unusual signals. Typically, these policies are updated based on rule‐based or deep learning algorithms to enhance detection performance. However, the insufficient number of labeled samples remains a challenge for real‐world deployment. In this article, an automated labeling method is presented that has shown high effectiveness, requiring minimal hardware resources, and applicable to IDS systems. Additionally, the approach utilizes transfer learning combined with attention mechanisms to boost the efficiency of abnormal signal detection. The results from the approach are compared with those of a reference model, illustrating an overall improvement of nearly 10% in our model's performance compared to the reference model. This underscores the effectiveness of automating rule adjustments for IDS, contributing significantly to reducing associated financial costs. The research addresses the challenges in deploying IDS in real‐world scenarios and provides a valuable contribution to enhancing cyber threat detection capabilities. A preprint has previously been published [11].

Abstract Exposure to ionizing radiation can induce genetic aberrations via unrepaired DNA strand breaks. To investigate quantitatively the dose–effect relationship at the molecular level, we irradiated dry pBR322 plasmid DNA with 3 MeV protons and assessed fragmentation yields at different radiation doses using long-read sequencing from Oxford Nanopore Technologies. This technology applied to a reference DNA model revealed dose-dependent fragmentation, as evidenced by read length distributions, showing no discernible radiation sensitivity in specific genetic sequences. In addition, we propose a method for directly measuring the single-strand break (SSB) yield. Furthermore, through a comparative study with a collection of previous works on dry DNA irradiation, we show that the irradiation protocol leads to biases in the definition of ionizing sources. We support this scenario by discussing the size distributions of nanopore sequencing reads in the light of Geant4 and Geant4-DNA simulation toolkit predictions. We show that integrating long-read sequencing technologies with advanced Monte Carlo simulations paves a promising path toward advancing our comprehension and prediction of radiation-induced DNA fragmentation.

Abstract Background Breastfeeding is the biological norm for feeding infants and young children. When mothers’ breastmilk is unavailable, donor human milk (DHM) from a human milk bank (HMB) becomes the next option for small vulnerable newborns. A comprehensive cost analysis is essential for understanding the investments needed to establish, operate, and scale up HMBs. This study aims to estimate and analyze such costs at the first facility established in Vietnam. Methods An activity-based costing ingredients (ABC-I) approach was employed, with the cost perspective from service provision agencies (specifically, the project conducted at Da Nang Hospital for Women and Children and Development Partners). Estimated financial costs, based on actual expenditures, were measured in 2023 local currency and then converted to 2023 US dollars (USD). We examined three scenarios: 1) direct start-up costs + indirect start-up costs + implementation costs, 2) direct start-up costs + implementation costs, and 3) capital costs + implementation costs over the 6.5 years of operation. Results The total start-up cost was USD 616,263, with total expenditure on direct activities at USD 228,131 and indirect activities at USD 388,132. Investment in equipment accounted for the largest proportion (USD 84,213). The monthly costs of Da Nang HMB were USD 25,217, 14,565, and 9,326, corresponding to scenarios 1, 2, and 3, respectively. Over HMB's 6.5 years of operation, on average, the unit costs were USD 166, USD 96, and USD 62 for DHM received and USD 201, USD 116, and USD 74 for pasteurized DHM meeting specified criteria in the corresponding scenarios. Unit costs were highest in the initial six months, decreased, and reached their lowest levels after a year. Then, the unit costs experienced an increase in late 2020 and early 2021. Conclusion Although the unit cost of DHM in Da Nang HMB is comparable to that in certain neighboring countries, intentional measures to reduce disposal rates, improve HMB efficiency, motivate more community-based donors, and establish an HMB service network should be implemented to lower costs.

Motors play a crucial role in production systems. However, not everything always goes smoothly, and motor failures are one of the common challenges in the production process. Misalignment of the drive shaft is a frequent motor fault caused by improper installation or damage to machine components. This study proposes the design of a monitoring and fault diagnosis model for DC motors, which includes: (i) a PID controller for motor speed control; (ii) a vibration signal acquisition unit; and (iii) a motor monitoring unit via Blynk and signal processing for fault diagnosis. In the model, motor faults are classified using a convolutional neural network (CNN) based on analog signals that have been transformed to the frequency domain and denoised. Experimental results demonstrate that classification using the convolutional neural network is highly accurate and stable.

Abstract Antimicrobial resistance (AMR) poses a serious threat to the clinical management of typhoid fever. AMR in Salmonella Typhi (S. Typhi) is commonly associated with the H58 lineage, a lineage that arose comparatively recently before becoming globally disseminated. To better understand when and how H58 emerged and became dominant, we performed detailed phylogenetic analyses on contemporary genome sequences from S. Typhi isolated in the period spanning the emergence. Our dataset, which contains the earliest described H58 S. Typhi organism, indicates that ancestral H58 organisms were already multi-drug resistant (MDR). These organisms emerged spontaneously in India in 1987 and became radially distributed throughout South Asia and then globally in the ensuing years. These early organisms were associated with a single long branch, possessing mutations associated with increased bile tolerance, suggesting that the first H58 organism was generated during chronic carriage. The subsequent use of fluoroquinolones led to several independent mutations in gyrA. The ability of H58 to acquire and maintain AMR genes continues to pose a threat, as extensively drug-resistant (XDR; MDR plus resistance to ciprofloxacin and third generation cephalosporins) variants, have emerged recently in this lineage. Understanding where and how H58 S. Typhi originated and became successful is key to understand how AMR drives successful lineages of bacterial pathogens. Additionally, these data can inform optimal targeting of typhoid conjugate vaccines (TCVs) for reducing the potential for emergence and the impact of new drug-resistant variants. Emphasis should also be placed upon the prospective identification and treatment of chronic carriers to prevent the emergence of new drug resistant variants with the ability to spread efficiently.

In Vietnam's rapidly expanding e-commerce landscape, there is a critical need for advanced tools that can effectively analyze customer feedback to boost satisfaction and loyalty. This paper introduces a two-step predictive framework merging deep learning and traditional machine learning to analyze Vietnamese e-commerce reviews. Utilizing a dataset of 10,021 reviews on Tiki, Shopee, Sendo, and Hasaki between 2015 and 2023, the framework first employs fine-tuned deep learning models like BERT and Bi-GRU to extract aspect-based sentiments from reviews, tailored for the nuances of the Vietnamese language. Subsequently, machine learning algorithms like XGBoost predict customer satisfaction by integrating sentiment analysis with e-commerce data such as product prices. Results show BERT and Bi-GRU yield over 70% sentiment accuracy, while XGBoost achieves 80%+ satisfaction prediction accuracy. This framework offers a potent solution for discerning customer sentiments and enhancing satisfaction in Vietnam's dynamic e-commerce landscape.

In this study, various environmentally friendly approaches such as microwave-assisted extraction (MAE), ultrasound-assisted extraction (UAE) and enzyme-ultrasound assisted extraction (EUAE) were investigated to simultaneously separate polysaccharides and polyphenols from mangosteen fruit peels (MPs). The yield of polysaccharides (PSY) and total phenolic content (TPC) were computed to compare the efficiency among methods. The chemical structure through Fourier-transform infrared (FTIR) spectrophotometry of the extracted polysaccharides as well as the antioxidant properties of the phenolic extracts were also determined. The result reveals that EUAE at the optimal condition (enzyme content of 6.6 U/g, enzymatic incubation time of 60 min, pH of 5.0 and ultrasonic duration of 30 min) provided much higher PSY (32.65%) and TPC of (107.8 mg gallic acid/ g of dry weight of sample) than MAE and UAE. The FTIR spectrum of the EUAE polysaccharides were remarkably distinct with those of MAE and UAE ones, indicating the differences in chemistry of the polysaccharides to be extracted. The findings of the study indicate that the polysaccharides and polyphenols from MPs by EUAE treatment can be used as natural ingredients in pharmaceutical and functional food formulations.

The Russian economic space is characterised by a significant differentiation in the levels of socio-economic development of the country's various regions, which manifests itself in natural, territorial, socio-cultural, economic, political and other aspects. The results of socio-economic differentiation are unique regional socio-economic systems, which necessitates the formation of individual approaches to managing their development. Therefore, management decisions made at the federal centre, as well as by regional authorities, affect the activities of economic entities in the regions and the population's level of well-being in different ways. Social security is an integral element of the high quality of life of the population and is largely the basis for improving the economic status of the region, increasing the value of human capital. Thus, it is necessary to develop methods and tools for ensuring the social safe development of regional socio-economic systems, considering the specific characteristics of each region. From these perspectives, we can discuss the stability and social performance of the regional economy. Despite a broad scientific background, the factors contributing to the development and the results of regional socio-economic systems, considering the need for social security, have not been examined. The present research aims to fill this gap by developing a management model for the social and safe development of Russia's regions, using the city of St Petersburg as the case study.

This paper provides a novel and applicable work that builds a real system for disinfecting the air and surfaces of the environment in a hospital room, with a non-contact measurement system for supporting contagious disease treatments in hospitals. The system is built on an intelligent mobile robot system that operates autonomously in a simulated real treatment room. The research team uses a new positioning algorithm. It is a combination of data from the Lidar sensor, encoder, and Extended Kalman filter. The program that applies segmentation and image feature extraction algorithms is developed to meet requirements of real-time environment mapping in the room. Control algorithms for moving and avoiding obstacles are also proposed. Next, techniques for collecting health data including patient identification, body temperature, and blood oxygen index via wireless sensor network are also mentioned in the article. Analysis and experimental results show qualified outcomes and promise. The main contribution of the paper can be listed as follows. • Design and build a new CEE-IMR, an intelligent mobile robot that can regconize patients, guide and lead them walking in hospitals, especially keep a safe distance avoiding contagious deseases. • A novel framework for controlling the robot is proposed. The robot can move flexible, avoid obstacles, etc. based on advanced control algorithms. A new control mechanism is also proposed. • Methods of collecting data and processing medical data to support either patients or doctors to improve the effecency in hospitals in contagious disease management.

Abstract This study aims to investigate the effect of organically modified zirconium dioxide nanoparticles (m-Zr[O.sub.2] NPs) on the mechanical properties, morphology, and weather stability of fluoroethylene/vinyl ether copolymer (FEVE). To [...]

The concept of graduate employability has gained great prominence in international education. However, there still exists a gap in sexual orientation discrimination in graduate employability among transgender and queer (TQ) international students. In our qualitative study investigating graduate employability of transgender and queer students graduating from Australian and Canadian institutions, we have interviewed 14 international graduates with transgender and queer identity regarding their perceptions of sexual orientation and recruitment discrimination at the workplaces. Utilizing intersectionality as a conceptual framework, we have studied employability-related problems that these marginalized students with their foreigner identities have experienced in the labor market. The findings will be around the social, cultural, and political impacts of Canadian and Australian working and recruitment environments on the varying extent of discrimination, namely local attitudes toward queer and transgender international graduates, the manifestation of antidiscrimination laws, and the extent to which employers value stereotypically male heterosexual personality traits. [The page range cited on the .pdf (p152-) is incorrect. The correct page range is p153-160.]

Accurate and robust pedestrian detection is fundamental for indoor robotic systems to navigate safely and seamlessly alongside humans in spatially constrained, unpredictable indoor environments. This paper presents a novel method, IRBGHR-PIXOR, a detection framework specifically engineered for pedestrian perception in indoor mobile robots. This novel approach employs an enhanced adaptation of the cutting-edge PIXOR model, integrating two pivotal augmentations: a remodeled convolutional backbone leveraging Inverted Residual Blocks (IRB) in unison with Gaussian Heatmap Regression (GHR), as well as a Modified Focal Loss (MFL) function to tackle data imbalance issues. The IRB component notably bolsters the network’s aptitude for processing intricate spatial representations generated from sparse 3D LiDAR scans. Meanwhile, integrating GHR further elevates accuracy by enabling precise localization of pedestrian subjects. This is achieved by modeling the probability distribution and predicting the central location of individuals in the point cloud data. Extensively evaluated on the large-scale JRDB dataset comprising intense scans from 16-beam Velodyne LiDAR sensors, IRBGHR-PIXOR accomplishes exceptional results, attaining 97.17% Average Precision (AP) at the 0.5 IOU threshold. Notably, this level of accuracy is achieved without significantly increasing model complexity. By enhancing algorithms to overcome challenges in confined indoor environments, this research paves the way for safe and effective deployment of autonomous technologies once encumbered by perceptual limitations in human-centered spaces. Nonetheless, evaluating performance in diverse edge cases and integration with complementary sensory cues promise continued progress. The developments contribute towards the vital capacity of reliable dynamic perception for next-generation robotic systems coexisting in human-centric environments.

The growth of last-mile delivery in developing countries is featuring the important contribution of crowdshippers. Therefore, exploring the factors affecting the intentions to stay in jobs of the crowdshippers is of great significance to maintain and sustainably develop the last-mile delivery service. Based on the Social Exchange Theory, we hypothesize that benefits (economic benefits, professional reputation, and job autonomy) and costs (risks and stress at work) have an effect on crowdshippers’ intentions to continue participating in last-mile delivery service in Vietnam. We also look for differences in motivations to stay in job between full-time and part-time workers. Survey data from 428 crowdshippers were analyzed using SmartPLS. The research findings indicate that economic benefits and work autonomy (benefits) have positive influences on intention to continue participating in crowd logistics, while risks and job stress (cost) have negative effects on the intention of crowsdhippers to continue their work. In addition, research findings also uncover that differences in motivation (perception of benefits and costs) to stay in job between full-time and part-time workers also affect their intentions. Although reputation did not show a significant impact on the crowdshippers’ intentions to continue participating in crowd logistics, it was discovered that reputation has a positive effect on full-time crowdshippers’ intentions to continue working. Based on the findings of this study, we recommend that businesses increase benefits and reduce costs in job to sustain and increase the continued participation in last-mile delivery.

Transmembrane protein 119 (TMEM119) is a recently identified microglia marker that is not expressed by other immune cells. Using CRISPR/Cas9 technology, we introduced enhanced green fluorescence protein (EGFP), into the H9 WA-09 human embryonic stem cell line, directly before the TMEM119 stop codon. Sanger sequencing confirmed successful insertion of the EGFP sequence. The newly created cell line expressed a normal morphology and karyotype, several pluripotency markers, and the ability to differentiate into all three germ layers. H9-TMEM119-EGFP can be used to provide a deeper understanding of the role of TMEM119 in microglia by monitoring its expression under different experimental conditions.

This study presents an analysis based on data collected via questionnaire, surveying Gen Z customers using food delivery applications in Vietnam. The purpose of the original research was to investigate factors influencing Gen Z customers' decision to continue using the applications. The data set presented in this paper includes 361 valid responses that were collected by convenience sampling method from Hanoi and Hochiminh City, which are the two most potential regions of e-commerce transactions in Vietnam. After being collected, sorted, and filtered, the data was calculated by SPSS 22 and AMOS 23 software to extract descriptive analysis, Cronbach's Alpha, and confirmatory factor analysis (CFA). The calculation results indicated that this data set ensures reliability, convergent, and discriminant validity, which can serve as a good reference for future studies.

Hand-raising gesture is one of the most popular signs of communication, whose frequency is related to the classroom’s atmosphere, the attractiveness of the subject, and the level of interactions between students and teachers. However, automatic hand-raising gesture detection and recognition remains still a challenging problem mainly due to low hand resolution, hand occlusion, various backgrounds, and viewpoint changes. While majority of the existing methods focus on static hand-raising posture detection, in this paper, we propose a framework for dynamic hand gesture recognition from classroom videos consisting of two main stages: hand posture detection and dynamic hand gesture recognition. In hand posture detection stage, we extend the previous work by adding relative position-aware in non-local network. After detecting the hand-raising posture on static pixels, which in turn accelerates the performance images, at the dynamic hand gesture recognition stage, we incorporate object detection and tracking to associate the hand-raising detection results at consecutive frames to supplement the missing detection due to the occlusion issue and obtain hand-raising gesture recognition at the event level. The experimental results show that the proposed method outperforms three benchmark models for static hand posture detection that are Faster-RCNN, Libra-RCNN, and Libra-RCNN+RDA on our dataset with the gains obtained for mAP being 7.68%, 5.76%, and 0.35% higher, respectively. In terms of event-level recognition results, the proposed method achieves the value of frame-wise accuracy, temporal_IoU, F1-score@0.3, and Levenshtein-score are 90.0%, 84.4%, 83.2%, and 84.3%, respectively. The code and dataset used in the paper will be made publicity available for research community.

Background: Appendiceal neuroendocrine tumors (NETs) rank as the third most frequent neoplasm affecting the appendix, originating from enterochromaffin cells. This study aims to evaluate the influence of various prognostic factors on the mortality rates of patients diagnosed with NETs of the appendix. Methods: Conducted retrospectively, the study involved 3346 patients, utilizing data sourced from the Surveillance, Epidemiology, and End Results (SEER) database. Our analysis centered on investigating demographic characteristics, clinical features, overall mortality (OM), and cancer-specific mortality (CSM) among the cohort. Variables showing a p-value < 0.1 in the univariate Cox regression were incorporated into the multivariate Cox regression analysis. A Hazard Ratio (HR) > 1 indicated an unfavorable prognosis. Results: In the multivariate analysis, higher OM and CSM were observed in males, older age groups, tumors with distant metastasis, poorly differentiated tumors, and those who underwent chemotherapy. Non-Hispanic Black individuals showed elevated mortality rates. Conclusion: Delayed diagnosis may contribute to the increased mortality in this community. Improved access to healthcare and treatment is crucial for addressing these disparities. Larger prospective studies are needed to pinpoint the underlying causes of elevated mortality in non-Hispanic Black populations, and randomized controlled trials (RCTs) are warranted to evaluate therapies for advanced-stage appendix NETs.

There is a growing recognition globally that care regarding lactation following a perinatal death needs to potentially offer the opportunity for maternal donation. This article discusses this experience and perspectives from a human milk bank (HMB) in Vietnam. This is a descriptive exploratory case study that has a long tradition in both the social and health sciences. Triangulated data collection involved a review of video data, interview data with the donor, and data review for the Da Nang HMB, a Center for Excellence in Breastfeeding. We found that although it is common for mothers in Vietnam to donate breastmilk to HMBs, it is less common for this to occur following perinatal loss. We offer a descriptive case study of the maternal loss of twins and a subsequent choice to donate for approximately 1 month to the Da Nang HMB, the first HMB in Vietnam. We discuss four reasons derived from this case regarding donation following perinatal loss. (1) A strong motivation to donate breastmilk when aware of the service, (2) donating breastmilk helped her deal with grief, (3) family members supported her through this tough time and supported her decision, and (4) health staff supported her decision. While human milk sharing (e.g., wet nursing) has been practiced in Vietnam, breastmilk donation from bereaved mothers has neither been discussed nor well-researched. Because maternal grief is complex and individual, deciding to donate breastmilk is a personal decision that needs to be supported, without creating guilt for those who do not wish to donate.

Summary: SCN2A protein-truncating variants (PTV) can result in neurological disorders such as autism spectrum disorder and intellectual disability, but they are less likely to cause epilepsy in comparison to missense variants. While in vitro studies showed PTV reduce action potential firing, consequences at in vivo network level remain elusive. Here, we generated a mouse model of Scn2a insufficiency using antisense oligonucleotides (Scn2a ASO mice), which recapitulated key clinical feature of SCN2A PTV disorders. Simultaneous two-photon Ca2+ imaging and electrocorticography (ECoG) in awake mice showed that spontaneous Ca2+ transients in somatosensory cortical neurons, as well as their pairwise co-activities were generally decreased in Scn2a ASO mice during spontaneous awake state and induced seizure state. The reduction of neuronal activities and paired co-activity are mechanisms associated with motor, social and cognitive deficits observed in our mouse model of severe Scn2a insufficiency, indicating these are likely mechanisms driving SCN2A PTV pathology.

BackgroundDonor human milk (DHM) is essential to the operation of human milk banks (HMB). This study examined characteristics and factors associated with higher volumes of DHM donation at the first HMB in Vietnam.MethodData from an online HMB monitoring system collected between February 2017 and July 2022 included demographic characteristics, child information, the timing of donation, and the volume of DHM. Higher volume is defined as equal to or greater than the median DHM volume per donor of 14.4 liters (L). Potential contributors to higher DHM volume were examined using the chi-square test in univariate and multivariable logistic regression analysis.ResultsDuring the 5.5-year operation, this HMB recruited 517 donors with an average age of 28.6 years. Approximately 60.9% of donors had a college or higher degree and 97.3% gave birth in Da Nang city. Of these donors, the prevalence of cesarean birth was 48.2%, preterm births was 40.2%, and 27.9% had babies with a birth weight of less than 1,500 g. There was a similar proportion of donors between the hospital (48.2%) and community (51.8%). On average, hospital donors started their donations 15 days after birth when their newborns were 33.9 weeks corrected age compared to 63 days and 47.7 weeks among community-based donors. The overall median volume of DHM per donor was 14.4 L over a period of 46 days. The amount and duration were higher in community-based donors (17.5 L in 72 days, 300 ml/day) than those in the hospital (8.4 L in 30 days, 258 ml/day). More than 37% of donors contacted the HMB themselves; the remainder were introduced by health professionals. Factors associated with higher volumes of DHM included higher education (OR: 1.77; 95% CI: 1.09, 2.87), having a full-term newborn (OR: 2.46; 95% CI: 1.46, 4.13), and community-based donors (OR: 2.15; 95% CI: 1.22, 3.78).ConclusionsMothers with higher education and from the community donate more breastmilk over a longer duration than those with lower education and from the hospital. Breastfeeding protection, promotion, and support should be offered to all mothers with specialized breastfeeding support for mothers of preterm and sick infants. This will ensure mothers have sufficient breastmilk for their newborns and potentially surplus breastmilk for donation.

Background: Routine placement of a drain for complicated appendicitis is one of the approaches to reduce postoperative complication. In a resource-limited hospital where radiologist-directed monitoring is not feasible, drainage is beneficial for complication follow-up. Nevertheless, emerging studies indicated drainage did not significantly impact on amelioration of patients’ outcomes. This study aimed to compare the postoperative results of pediatric patients with or without drainage after laparoscopic appendectomy for complicated appendicitis at a rural children's hospital. Methods: We performed a quasi-randomized recruitment of pediatric patients who underwent laparoscopic treatment for complicated appendicitis with or without drainage. Several parameters, including operative time, lengths of stay, complication rate, duration of antibiotic administration, and hospital charges were monitored for assessment. Results: From 05/2020 to 05/2022, there were 184 cases of recruited pediatric patients who underwent laparoscopic appendectomy for complicated appendicitis, with 91 patients in drain-placing group (49.5%). Demographic baseline and pre-surgery symptoms did not indicate any differences between two groups. Data of postoperative complications including intra-abdominal abscesses and wound infections showed no significance in discrepancy. Notably, patients of drainage group experienced longer length of stay than non-drainage group, with average of 8.5 days and 7.5 days, respectively (p

Plant segmentation is a critical task in precision agriculture as related to crop management and weed treatment. Plants can exhibit very large scale changes, which presents great challenge for accurate crop/weed segmentation. Recent works have shown that multi-scale features are useful to segment objects with different scales. In this work, we propose a Dense Multi-scale Convolutional Network (DMSCN) for pixel-wise crop/weed segmentation. Our network has an encoder-decoder structure. The encoder comprises of a Dense Convolutional Network (DCN) and a Dense Multi-Scale Atrous Pooling (DMSAP) module. DCN is composed of standard and atrous convolutions with dense connections. The architecture of DCN allows the encoder to increase the density of feature maps while avoiding signal decimation due to the dimension reduction. The proposed DMSAP connects a set of standard and atrous convolutional layers with different dilation rates in a densely cascaded manner. DMSAP is able to capture features with dense scale sampling and large receptive field. A simple yet effective decoder is used to refine the segmentation results by combining high and low-level features of the encoder. Extensive experiments are performed on four crop/weed datasets. One of these datasets was collected and annotated by us. We conduct an ablation study to show the advantages of different modules of DMSCN. The comparative study demonstrates the advantages of our model compared with the previous methods in terms of accuracy and complexity.