The deployment of many advanced pedestrian detection applications is largely hindered by the high computational cost of deep convolutional neural networks (CNNs). In this paper, we propose a two-step pruning method to design a lightweight pedestrian detection network in railway scenes. The first step is feature pyramid network (FPN) pruning, which utilizes the characteristic of pedestrian in railway scenes and the FPN structure in YOLOv3. The second step is regular channel pruning, which utilizes network slimming knowledge and is an accelerator-friendly pruning strategy. Our two-step pruning method gives about 88% reduction in parameters and about 74% reduction in computing complexity with comparable detection accuracy.

Pedestrian detection is a classic problem in computer vision, which has an essential impact on the safety of urban autonomous driving. Although significant improvement has been made in pedestrian detection recently, small-scale pedestrian detection is still challenging. To effectively tackle this issue, a multi-scale pedestrian detector based on self-attention mechanism and adaptive spatial feature fusion is proposed in this paper. In order to better extract global information, the spatial attention mechanism asymmetric pyramid non-local block (APNB) module is applied. To achieve scale-invariance detection, multiple detection branches are designed, which include a high-resolution detection branch and a lowresolution detection branch. In integrating multi-scale features, the adaptively spatial feature fusion (ASFF) method is employed, which can solve the problem of feature inconsistency across different scales. Experimental results show that the proposed method obtains competitive performance on Caltech and CityPersons datasets.This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

No abstract

A population of Echinochloa crus-galli (L.) P. Beauv obtained from direct-seeding rice fields in Jiangxi Province, China, exhibited high resistance levels (13.5-fold) to the acetyl-CoA carboxylase (ACCase)-inhibiting herbicide cyhalofop-butyl. Compared with the susceptible (S) population, this resistant (R) population evolved a cross-resistance to aryloxyphenoxypropionates (APPs) herbicides metamifop (2.9-fold) and fenoxapro-p-ethyl (4.1-fold), cyclohexanediones (CHDs) herbicide clethodim (4.7-fold), phenyl pyrazoline (DEN) herbicide pinoxaden (6.4-fold), and evolved multiple-resistance to acetolactate synthase (ALS)-inhibiting herbicide penoxsulam (3.6-fold), and auxin mimic herbicides quinclorac (>34.7-fold) and florpyrauxifen-benzyl (2.4-fold). ACCase gene sequencing did not reveal the existence of any known mutation point conferring with herbicide resistance. In addition, three metabolic inhibitors—one glutathione—S-transferase (GST) inhibitor (NBD-Cl), and two cytochrome P450 inhibitors (malathion and PBO)—did not reverse the cyhalofop-butyl resistance. Furthermore, enhanced metabolic rates of more than 60% 24 h after treatment with the active compound cyhalofop acid was observed in R plants compared to S plants. Hence, enhanced metabolism activity endows a non-target-site resistance to cyhalofop-butyl in the R population of E. crus-galli. Future research will be required to determine what metabolizing enzyme genes are responsible for cyhalofop-butyl resistance in E. crus-galli.

A Gram-stain-negative, rod-shaped bacterium, strain F01T, was isolated from leaves of Tamarix chinensis Lour. The isolate grew optimally at 30 °C, at pH 7.0 and with 5.0 % (w/v) NaCl, and showed a high tolerance to manganese, lead, nickel, ferrous ions and copper ions. The major fatty acids were C18 : 1ω7c and C16 : 0, and the predominant respiratory quinone was Q-9. Polar lipids were dominated by diphosphatidylglycerol, phosphatidylethanolamine, phosphatidylglycerol, unidentified aminoglycolipids and phospholipids. The DNA G+C content was 65.8 %. Based on multilocus phylogenetic analysis, strain F01T belonged to the genus Salinicola, with highest 16S rRNA gene sequence similarity to Salinicola peritrichatus CGMCC 1.12381T (97.7 %). The level of DNA-DNA hybridization between strain F01T and closely related Salinicola strains was well below 70 %. According to the phenotypic, genetic and chemotaxonomic data, strain F01T is considered to represent a novel species in the genus Salinicola, for which the name Salinicola tamaricis sp. nov. is proposed. The type strain is F01T (=CCTCC AB 2015304T=KCTC 42855T).