Search

Recently, significant improvements on face detection have been achieved by convolutional neural networks. However, the speed of a CNN-based face detector is hampered by the large computational complexity and the huge number of parameters. It is still challenging to achieve real-time face detection as well as maintain high accuracy. In this work, we introduce a single-stage face detector with an extremely lightweight CNN to achieve fast and accurate detection. Specifically, our method has a structure to integrate the network with the image pyramid for fully utilizing the calculated features. Benefiting from weight sharing, the network size still can keep small. We also analysis the detection capability of anchors with various scales, and reserve the most effective anchors in our model. Besides, to avoid the too difficult training samples which the small network can’t learn, each ground truth face is assigned with a 0-or-1 weight during training. When tested on WIDERFACE and FDDB, it outperforms existing lightweight face detectors on accuracy with the smallest model size. The outstanding detection performance and lightweight model size signify its effectiveness and practicability.

Spiking neural networks (SNNs) are currently one of the most successful approaches to model the behavior and learning potential of the brain. Recently, they have obtained marvelous research interest thanks to their event-driven and energy-efficient characteristics. While difficult to directly train SNNs from scratch because of their non-differentiable spike operations, many works have focused on converting a trained DNN to the target SNN. However, there is no efficient method to convert the deformable convolutional layer which is frequently used in many applications. The deformable convolution layer enables deformation of the convolutional sampling grid by adding offsets to the regular sampling locations, which enhances the geometric transformation modeling capability of CNNs. In this work, we propose a novel deformable spiking CNN, which can successfully convert DNNs with deformable convolution layers to SNNs with much shorter simulation time and have low latency during inference while maintaining high accuracy. To be specific, we design an effective method dedicated for deformable convolution layers to be converted. By treating the offset prediction module as an embedded SNN, we calculate the spiking offsets multi times and use the average values as the final offsets for deformable convolution. We also propose a stage-wise DNN-SNN conversion strategy to further reduce the conversion error. We divide the network into several stages and convert each stage sequentially with retraining to diminish the difference between the source DNN and the target SNN as much as possible. The experiments on CIFAR-10 and CIFAR-100 datasets show that our method surpasses the state-of-the-art works both in conversion accuracy and inference latency.

Background Insecticide resistance is a substantial problem in controlling agricultural and medical pests. Detecting target site mutations is crucial to manage insecticide resistance. Though PCR-based methods have been widely used in this field, they are time-consuming and inefficient, and typically have a high false positive rate. Acetylcholinesterases (Ace) is the neural target of the widely used organophosphate (OP) and carbamate insecticides. However, there is not any software available to detect insecticide resistance associated mutations in RNA-Seq data at present. Results A computational pipeline ACE was developed to detect resistance mutations of ace in insect RNA-Seq data. Known ace resistance mutations were collected and used as a reference. We constructed a Web server for ACE, and the standalone software in both Linux and Windows versions is available for download. ACE was used to analyse 971 RNA-Seq data from 136 studies in 7 insect pests. The mutation frequency of each RNA-Seq dataset was calculated. The results indicated that the resistance frequency was 30%–44% in an eastern Ugandan Anopheles population, thus suggesting this resistance-conferring mutation has reached high frequency in these mosquitoes in Uganda. Analyses of RNA-Seq data from the diamondback moth Plutella xylostella indicated that the G227A mutation was positively related with resistance levels to organophosphate or carbamate insecticides. The wasp Nasonia vitripennis had a low frequency of resistant reads (

Insect pests reduce yield and cause economic losses, which are major problems in agriculture. Parasitic wasps are the natural enemies of many agricultural pests and thus have been widely used as biological control agents. Plants, phytophagous insects and parasitic wasps form a tritrophic food chain. Understanding the interactions in this tritrophic system should be helpful for developing parasitic wasps for pest control and deciphering the mechanisms of parasitism. However, the genomic resources for this tritrophic system are not well organized. Here, we describe the WaspBase, a new database that contains 573 transcriptomes of 35 parasitic wasps and the genomes of 12 parasitic wasps, 5 insect hosts and 8 plants. In addition, we identified long non-coding RNA, untranslated regions and 25 widely studied gene families from the genome and transcriptome data of these species. WaspBase provides conventional web services such as Basic Local Alignment Search Tool, search and download, together with several widely used tools such as profile hidden Markov model, Multiple Alignment using Fast Fourier Transform, automated alignment trimming and JBrowse. We also present a collection of active researchers in the field of parasitic wasps, which should be useful for constructing scientific networks in this field.

In multi-label classification, high-dimensional and sparse binary label vectors usually make existing multi-label classifiers perform unsatisfactorily, which induces a group of label compression coding (LCC) techniques particularly. So far, several linear LCC methods have been introduced via considering linear relations among labels. In this paper, we extend traditional three-layer auto-encoder to construct a five-layer one (i.e., five-layer symmetrical neural network), and then apply the training principle in extreme learning machine to determine all network weights. Therefore, a non-linear LCC approach is proposed to capture non-linear relations of labels, where the first three-layer network is regarded as a encoder and the last two layers act as a decoder. The experimental results on three benchmark data sets show that our proposed method performs better than four existing linear LCC methods according to five performance measures.