Search

Lightweight deep convolutional neural networks (LDCNNs) are vital components of mobile intelligence, particularly in mobile vision. Although various heavy networks with increasingly deeper and wider have continuously broken accuracy records since 2012, with the spring of terminals and mobile devices, neural networks that can match them have become a core role in practical applications. In this review, we focus on several representative lightweight Deep Convolutional Neural Networks (DCNN) technologies that hold significant potential for advancing the field. More than 190 references screened out in terms of architecture design and model compression, in which over 50 representative ones are emphasized from the perspectives of methods, performance, advantages, and drawbacks, as well as underlying framework support and benchmark datasets. With a comprehensive analysis, we put forward some existing problems and offer prospects of lightweight DCNN for future development. [ABSTRACT FROM AUTHOR]

Lead, a toxic heavy metal present in soil, hampers biological activities and affects the metabolism of plants, animals, and human beings. Its higher concentration may disturb the various physio-chemical processes, which result in stunted and poor plant growth. An interactive approach of plant growth promoting rhizobacteria (PGPR) and L-tryptophan can be used to mitigate the lethal effects of lead. A pot experiment was conducted, and two weeks before sowing, the level of lead (300 mg kg−1) was maintained by spiking the PbCl2 salt. Pseudomonas fluorescens and L-tryptophan were applied individually as well as in combination to segregate the effect of both in contaminated soil under a completely Randomized Design (CRD). Statistical analysis revealed that plant growth was significantly reduced up to 22% due to lead contamination. However, the interactive approach of PGPR and L-tryptophan significantly improved the plant growth, physiology, and yield with relative productive index (RPI) under a lead-stressed environment. Moreover, integrated use of PGPR and L-tryptophan demonstrated a considerable increase (22%) in lead removal efficiency (LRE) by improving bioconcentration factor (BCF) and translocation factor (TF) for shoot without increasing the lead concentration in achenes. The reduced lead concentration in achene was due to its immobilization in shoot and root by negatively charged particles and improved the lead sequestration in vegetative parts which abridged the translocation of lead into achenes. [ABSTRACT FROM AUTHOR]

Direct borohydride fuel cell (DBFC) has been widely used in portable and mobile power supply because of its advantages of high theoretical cell voltage, considerable specific capacity and green environmental friendliness. Anode catalyst is the key material to improve DBFC performance and fuel efficiency. In the past, noble metals have been used as anode catalysts for DBFC batteries, but the high cost and limited resources of noble metals have limited the further development of the DBFC. At present, transition metal catalysts have received increasing attention due to their low cost and catalytic activity comparable to that of noble metals. On the basis of elucidating the working principle of DBFC, the research achievements of transition metal catalysts for DBFC in the past 20 years are reviewed in this paper. Through a large number of literature studies, it is found that the research results related to Ni-based, Co-based and their composite transition metal catalysts are the most abundant. It is also found that transition metal catalysts with nanometer size and loose structure show good catalytic activity. Therefore, Ni-based, Co-based and their composite transition metal catalysts were reviewed in detail. We believe that with the deepening of research, the proportion of transition metals in DBFC anode catalysts is increasing, which has a broader research space and application prospect. [ABSTRACT FROM AUTHOR]

Farmland has become a significant contributor to greenhouse gas (GHG) emissions, and research has shown that the addition of straw or biochar may be a viable method for mitigating these emissions. However, there is a lack of understanding regarding the comparative biotic and abiotic effects of straw and biochar amendments on GHG emissions. To address this knowledge gap, we conducted a meta-analysis of 100 published papers to quantify the impact of straw and biochar application on GHG emissions. Our findings indicate that straw application significantly increased CO2 and CH4 emissions from agricultural ecosystems by 46.2% and 113.5%, respectively, but did not have a significant effect on N2O emissions. Conversely, biochar amendment significantly reduced CO2, CH4, and N2O emissions by an average of 11.0%, 31.7%, and 22.8%, respectively. We also found that straw and biochar amendments increased soil pH, soil organic carbon (SOC), and C/N ratio, and there were significant differences between them. Moreover, straw application significantly increased the microbial biomass carbon (MBC) content and microbial quotient by 37.1% and 20.1%, respectively, while biochar application increased the MBC content by 25.0% without a significant effect on the microbial quotient. Furthermore, both straw and biochar applications promoted the nitrification process and increased the abundance of ammonia-oxidizing bacteria (AOB) by 50.7% with straw and by 57.5% and 75.1% with biochar for ammonia-oxidizing archaea (AOA) and AOB, respectively. The denitrification process was also stimulated by straw or biochar amendment, resulting in an increase in the abundance of nirK by 22.9% and 16.8%, respectively. Biochar amendment additionally increased the abundance of nosZ by 29.4%, indicating that the main reason for reducing N2O emissions through biochar application is the conversion of NO3−-N to N2. Thus, compared to straw application, biochar application is a more effective method for reducing greenhouse gas emissions. [ABSTRACT FROM AUTHOR]

Herein, we reported an efficient ultrasound-assisted method to synthesize the nanocomposite of graphitized and carboxylated multi-walled carbon nanotubes and chitosan (GCMCN@CTS), which was used to fabricate the GCMCN@CTS/GCE sensor for the detection of niclosamide (NA). GCMCN with graphitization and carboxylation possessed excellent electrical conductivity and high specific surface area. Graphitization enhanced the electrical conductivity of carbon nanotubes, and carboxylation improved the dispersion degree of carbon nanotubes. CTS with good biocompatibility possessed strong adsorption ability and superior film-forming property. Moreover, the introduction of CTS facilitated the uniform dispersion of GCMCN, and GCMCN with excellent electrical conductivity could make up for the poor electrical conductivity of CTS. Benefitting from the functional collaboration of GCMCN and CTS, an ultra-low detection limit of 1.4 nM could be achieved at the fabricated GCMCN@CTS/GCE sensor in linear NA concentration range of 0.01–10 μM. Moreover, the proposed sensor showed satisfactory reproducibility, repeatability, and anti-interference performance. When used for NA detection in tap water and lake water samples, the GCMCN@CTS/GCE sensor exhibited gratifying recoveries of 96.0–103.8% with RSD values of 2.0–4.7%. [ABSTRACT FROM AUTHOR]

There is an increasing demand for female fertility preservation. Cryopreservation of ovarian cortex tissue by means of vitrification can be done ad-hoc and for pre-pubertal individuals. Obtaining a homogeneous distribution of protective agents in tissues is one of the major hurdles for successful preservation. Therefore, to rationally design vitrification strategies for tissues, it is needed to determine permeation kinetics of cryoprotective agents; to ensure homogeneous distribution while minimizing exposure time and toxicity effects. In this study, Fourier transform infrared spectroscopy (FTIR) was used to monitor diffusion of different components into porcine ovarian cortex tissue. Water fluxes and permeation kinetics of dimethyl sulfoxide (DMSO), glycerol (GLY), ethylene glycol (EG), and propylene glycol (PG) were investigated. Diffusion coefficients derived from FTIR data, were corroborated with differential scanning calorimetry and osmometer measurements. FTIR allowed real-time spectral fingerprinting of tissue during loading with mixtures of protective agents, while discriminating between different components and water. Exposure to vitrification solutions was found to cause drastic initial weight losses, which could be correlated with spectral features. Use of heavy water allowed distinguishing water fluxes associated with dehydration and permeation, both of which were found to precede permeation of cryoprotective agents. Overall, DMSO and EG were found to permeate faster than GLY and PG. In mixtures, however, solutes behave differently. The non-invasive spectroscopic method described here to study permeation of vitrification solution components into ovarian tissue can be applied to many other types of engineered constructs, tissues, and possibly organs. [ABSTRACT FROM AUTHOR]