An Attention Mechanism-Improved YOLOv7 Object Detection Algorithm for Hemp Duck Count Estimation

Abstract

Stocking density presents a key factor affecting livestock and poultry production on a large scale as well as animal welfare. However, the current manual counting method used in the hemp duck breeding industry is inefficient, costly in labor, less accurate, and prone to double counting and omission. In this regard, this paper uses deep learning algorithms to achieve real-time monitoring of the number of dense hemp duck flocks and to promote the development of the intelligent farming industry. We constructed a new large-scale hemp duck object detection image dataset, which contains 1500 hemp duck object detection full-body frame labeling and head-only frame labeling. In addition, this paper proposes an improved attention mechanism YOLOv7 algorithm, CBAM-YOLOv7, adding three CBAM modules to the backbone network of YOLOv7 to improve the network’s ability to extract features and introducing SE-YOLOv7 and ECA-YOLOv7 for comparison experiments. The experimental results show that CBAM-YOLOv7 had higher precision, and the recall, mAP@0.5, and mAP@0.5:0.95 were slightly improved. The evaluation index value of CBAM-YOLOv7 improved more than those of SE-YOLOv7 and ECA-YOLOv7. In addition, we also conducted a comparison test between the two labeling methods and found that the head-only labeling method led to the loss of a high volume of feature information, and the full-body frame labeling method demonstrated a better detection effect. The results of the algorithm performance evaluation show that the intelligent hemp duck counting method proposed in this paper is feasible and can promote the development of smart reliable automated duck counting.

Research on surface garbage collection device based on YOLOv7

“…YOLOv7 [7] is similar to YOLOv5 on the whole, with the following major improvements: internal components and label allocation ideas in the network structure will be reintroduced into the network architecture; The cross-network search strategy of YOLOv5 and the matching strategy of YOLOX are applied to the label allocation strategy. A new efficient network architecture ELAN is proposed.…”

Tea leaf disease detection and identification based on YOLOv7 (YOLO-T)

“…The method of compound scaling allows for the preservation of the model's starting attributes and, consequently, the best structure. Then the model concentrates on several trainable optimization modules and techniques known as "bag-of-freebies" (BoF) 27,36 YOLOv7 is not confined to a single head, as it was inspired by deep supervision, a common training strategy for deep neural networks. It has several heads to accomplish anything it desires.…”

“…Traditional label assignment uses the ground truth directly to create hard labels based on preset criteria. Reliable soft labels, conversely, use calculation and optimization methods that consider both the ground truth and the quality and distribution of prediction output 27,36 . Figure 5 shows the overview of the network architecture diagram of YOLOv7.…”

“…This indicates that the batch's normalized mean and variance are added to the deviation and weight of the convolution layer during the inference step, (b) Using the addition and multiplication technique of knowledge acquisition in YOLO-R in combination with the convolution feature map, it can be standardized into vectors by pre-computation in the inference stage to be combined with the deviation and weight of the previous or subsequent convolution layer, and (c) finally, real-time object detection can considerably boost the detection accuracy without affecting the computational cost, so that the speed and precision in the range of 5-160 FPS surpass all known object detectors, enabling rapid response and accurate prediction of object detection 27,36 .…”

“…The upgraded YOLO (YOLOv7) variant has caught many machine learning and data modelling scientists. Several researchers have employed for various object-detecting phenomena, such as video object tracking 29 , object detection for Hemp duck count estimation 36 , object detection of maritime UAV images 37 , ship detection from satellite images 38 , defect detection in different materials [39][40][41] , vehicle tracking 42 , as well as in healthcare 43,44 . Gallo et al 45 applied the YOLOv7 model on a dataset of Chicory plants to identify weeds.…”