Insect pests strongly affect crop growth and value globally. Fast and precise pest detection and counting are crucial measures in the management and mitigation of pest infestations. In this area, deep learning technologies have come to represent the method with the most potential. However, for small-sized crop pests, recent deep-learning-based detection attempts have not accomplished accurate recognition and detection due to the challenges posed by feature extraction and positive and negative sample selection. Therefore, to overcome these limitations, we first designed a co-ordinate-attention-based feature pyramid network, termed CAFPN, to extract the salient visual features that distinguish small insects from each other. Subsequently, in the network training stage, a dynamic sample selection strategy using positive and negative weight functions, which considers both high classification scores and precise localization, was introduced. Finally, several experiments were conducted on our constructed large-scale crop pest datasets, the AgriPest 21 dataset and the IP102 dateset, achieving accuracy scores of 77.2% and 29.8% for mAP (mean average precision), demonstrating promising detection results when compared to other detectors.

A serious outbreak of agricultural pests results in a great loss of corn production. Therefore, accurate and robust corn pest detection is important during the early warning, which can achieve the prevention of the damage caused by corn pests. To obtain an accurate detection of corn pests, a new method based on a convolutional neural network is introduced in this paper. Firstly, a large-scale corn pest dataset has been constructed which includes 7741 corn pest images with 10 classes. Secondly, a deep residual network with deformable convolution has been introduced to obtain the features of the corn pest images. To address the detection task of multi-scale corn pests, an attention-based multi-scale feature pyramid network has been developed. Finally, we combined the proposed modules with a two-stage detector into a single network, which achieves the identification and localization of corn pests in an image. Experimental results on the corn pest dataset demonstrate that the proposed method has good performance compared with other methods. Specifically, the proposed method achieves 70.1% mean Average Precision (mAP) and 74.3% Recall at the speed of 17.0 frames per second (FPS), which balances the accuracy and efficiency.