Object Detection With Deep Learning: A Review.

Due to object detection's close relationship with video analysis and image understanding, it has attracted much research attention in recent years. Traditional object detection methods are built on handcrafted features and shallow trainable architectures. Their performance easily stagnates by constructing complex ensembles that combine multiple low-level image features with high-level context from object detectors and scene classifiers. With the rapid development in deep learning, more powerful tools, which are able to learn semantic, high-level, deeper features, are introduced to address the problems existing in traditional architectures. These models behave differently in network architecture, training strategy, and optimization function. In this paper, we provide a review of deep learning-based object detection frameworks. Our review begins with a brief introduction on the history of deep learning and its representative tool, namely, the convolutional neural network. Then, we focus on typical generic object detection architectures along with some modifications and useful tricks to improve detection performance further. As distinct specific detection tasks exhibit different characteristics, we also briefly survey several specific tasks, including salient object detection, face detection, and pedestrian detection. Experimental analyses are also provided to compare various methods and draw some meaningful conclusions. Finally, several promising directions and tasks are provided to serve as guidelines for future work in both object detection and relevant neural network-based learning systems.

Metagenome-Assembled Genome Sequence of Rhodopseudomonas palustris Strain ELI 1980, Commercialized as a Biostimulant.

We report here the draft genome sequence of strain ELI 1980 of Rhodopseudomonas palustris, commercialized as a biostimulant for agriculture. The genome was reconstructed from the metagenome of a commercial product containing this strain as its major component.

Impact of tobramycin on the performance of microbial fuel cell.

BACKGROUND: The release of antibiotics into aquatic environments has made the treatment of wastewater containing antibiotics a world-wide public health problem. The ability of microbial fuel cells (MFCs) to harvest electricity from organic waste and renewable biomass is attracting increased interest in wastewater treatment. In this paper we investigated the bioelectrochemical response of an electroactive mixed-culture biofilm in MFC to different tobramycin concentrations. RESULTS: The electroactive biofilms showed a high degree of robustness against tobramycin at the level of µg/L. The current generation responses of the biofilms were affected by the presence of tobramycin. The inhibition ratio of the MFC increased exponentially with the tobramycin concentrations in the range of 0.1-1.9 g/L. The bacterial communities of the biofilms vary with the concentrations of tobramycin, the equilibrium of which is critical for the stability of electroactive biofilms based-MFC. CONCLUSIONS: Experimental results demonstrate that the electroactive biofilm-based MFC is robust against antibiotics at the level of µg/L, but sensitive to changes in antibiotic concentration at the level of g/L. These results could provide significant information about the effects of antibiotics on the performance MFC as a waste-treatment technology.

Enhanced performance and mechanism study of microbial electrolysis cells using Fe nanoparticle-decorated anodes.

Anode properties are critical for the performance of microbial electrolysis cells (MECs). In the present study, Fe nanoparticle-modified graphite disks were used as anodes to investigate the effects of nanoparticles on the performance of Shewanella oneidensis MR-1 in MECs. Results demonstrated that the average current densities produced with Fe nanoparticle-decorated anodes up to 5.89-fold higher than plain graphite anodes. Whole genome microarray analysis of the gene expression showed that genes encoding biofilm formation were significantly up-regulated as a response to nanoparticle-decorated anodes. Increased expression of genes related to nanowires, flavins, and c-type cytochromes indicates that enhanced mechanisms of electron transfer to the anode may also have contributed to the observed increases in current density. The majority of the remaining differentially expressed genes associated with electron transport and anaerobic metabolism demonstrate a systemic response to increased power loads.

Nanoparticle decorated anodes for enhanced current generation in microbial electrochemical cells.

The development of highly efficient anode materials is critical for enhancing the current output of microbial electrochemical cells. In this study, Au and Pd nanoparticle decorated graphite anodes were developed and evaluated in a newly designed multi-anode microbial electrolysis cell (MEC). The anodes decorated with Au nanoparticles produced current densities up to 20-fold higher than plain graphite anodes by Shewanella oneidensis MR-1, while those of Pd-decorated anodes with similar morphologies produced 50-150% higher than the control. Significant positive linear regression was obtained between the current density and the particle size (average Feret's diameter and average area), while the circularity of the particles showed negative correlation with current densities. On the contrary, no significant correlation was evident between the current density and the particle density based on area fraction and particle counts. These results demonstrated that nano-decoration can greatly enhance the performance of microbial anodes, while the chemical composition, size and shape of the nanoparticles determined the extent of the enhancement.

Electricity generation from polyalcohols in single-chamber microbial fuel cells.

Six polyalcohols derived from lignocellulosic carbohydrates were investigated as carbon sources for electricity generation in single-chamber mediator-less microbial fuel cells (MFCs) for the first time. Electricity was directly generated from all polyalcohols tested, including pentitols (xylitol, arabitol, and ribitol) and hexitols (galactitol, mannitol, and sorbitol). Bacterial cultures initially enriched using acetate could be adapted to these substrates with varied adaptation times. The resultant maximum power density ranged from 1490+/-160 mW/m(2) to 2650+/-10 mW/m(2) at current densities between 0.58 mA/cm(2) and 0.78 mA/cm(2). Galactitol generated the highest maximum power density, while mannitol resulted in the lowest one. The estimated maximum voltage output at an external resistance of 120 Omega ranged between 0.24 V and 0.34 V with half saturation kinetic constants varied from 298 mg/L to 753 mg/L. The removal of chemical oxygen demand (COD) was above 91% for all polyalcohols except sorbitol (71%). Denaturing gradient gel electrophoresis (DGGE) of PCR-amplified 16S rRNA gene segments of the anode biofilms showed the influence of substrates (polyalcohols) on the anode microbial populations.

Effect of nitrate on the performance of single chamber air cathode microbial fuel cells.

The effect of nitrate on the performance of a single chamber air cathode MFC system and the denitrification activity in the system were investigated. The maximum voltage output was not affected by 8.0mM nitrate in the medium solution at higher external resistance (270-1000Omega), but affected at lower resistance (150Omega) possibly due to the low organic carbon availability. The Coulombic efficiency was greatly affected by the nitrate concentration possibly due to the competition between the electricity generation and denitrification processes. Over 84-90% of nitrate (0.8-8.0mM) was removed from the single chamber MFCs in less than 8h in the first batch. After 4-month operation, over 85% of nitrate (8.0mM) was removed in 1h after the MFC was continuously fed with a medium solution containing nitrate. Only a small amount of nitrite (<0.01mM) was detected during the denitrification process. The similar denitrification activity observed at different external resistances (1000 and 270Omega) and open circuit mode indicates that the denitrification was not significantly affected by the electricity generation process. No electricity was generated when the MFC fed with 8.0mM nitrate was moved to a glove box (no oxygen), indicating that the bacteria on the cathode did not involve in accepting electrons from the circuit to reduce the nitrate. Denaturing Gradient Gel Electrophoresis (DGGE) profiles demonstrate a similar bacterial community composition on the electrodes and in the solution but with different dominant species.

[Field efficacy of emulsifiable suspensions of Beauveria bassiana conidia for control of Myzus persicae population on cabbage].

The conidial powder of Beauveria bassiana SG8702 produced using a diphasic fermentation technology was formulated into emulsifiable suspension I (EI-I, > 10(10) conidia.ml-1) and ES-II, which resulted from adding 1% (W/V) of imidacloprid 10% WP to the ES-I. Three dilutions (10(7), 10(6), and 10(5) conidia.ml-1 of each were sprayed onto cabbage plants for control of the green peach aphid Myzus persicae population in field plots in Kunming, Yunnan during July 2001. Each dilution and a water spray as control were replicated three times. During a 28 d period of sampling at 3 or 4 d intervals, spray of the ES-II at 10(7) conidia.ml-1 controlled the increase of the aphid population by > 90% consistently from day 7 after spray. The efficacy for the ES-I at the same concentration reached 85% on day 7 after spray, exceeded 70% in the following two weeks, but declined to 64.4% on day 24 and 52.6% on day 28. At the lower concentrations sprayed, both suspensions significantly suppressed the increase of M. persicae population, but the ES-II resulted in higher efficacy than the ES-I. Local summer weather with mild temperature and frequent rainy days was favorable to the use of the formulations for aphid control.

[Time-specific infection rate of Beauveria bassiana on Myzus persicae after topical inoculation of conidial suspension].

Conidial suspension(5 x 10(6) conidia.ml-1) of the entomopathogenic fungus, Beauveria bassiana BBSG8702 was used to inoculate newly emerged apterous adults of the green peach aphid (Myzus persicae) on detached cabbage leaves in order to measure time-specific infection rate of the fungal agent. The aphids of conidia attached were disinfected using 0.2% chlorothalonil solution at intervals of 8-56 h after inoculation and then maintained at 10 and 20 degrees C for daily observation. There was a significant difference (P < 0.05) in mortality attributed to the fungal infection among the time intervals within 56 h at 10 degrees C after inoculation or within 40 h at 20 degrees C. Compared with the control (not treated with 0.2% chlorothalonil after inoculation), time-specific infection rate at 10 degrees C was estimated as 22.9%, 48.8%, 64.9%, 80.4%, 72.7%, and 98.3% within 8 h, 16 h, 24 h, 32 h, 40 h, and 56 h after inoculation, respectively. The same estimates at 20 degrees C were 31.6%, 48.8%, 58.6%, 86.9%, 97.2%, and 98.7%, respectively. The results indicate that the first 24 h after inoculation was crucial to the infection of B. bassiana to M. persicae. Effective infection with the first 24 h accumulated to 59-65% at 10-20 degrees C.