High-resolution processing and sigmoid fusion modules for efficient detection of small objects in an embedded system

Abstract

Recent advances in deep learning realized accurate, robust detection of various types of objects including pedestrians on the road, defect regions in the manufacturing process, human organs in medical images, and dangerous materials passing through the airport checkpoint. Specifically, small object detection implemented as an embedded system is gaining increasing attention for autonomous vehicles, drone reconnaissance, and microscopic imagery. In this paper, we present a light-weight small object detection model using two plug-in modules: (1) high-resolution processing module (HRPM ) and (2) sigmoid fusion module (SFM). The HRPM efficiently learns multi-scale features of small objects using a significantly reduced computational cost, and the SFM alleviates mis-classification errors due to spatial noise by adjusting weights on the lost small object information. Combination of HRPM and SFM significantly improved the detection accuracy with a low amount of computation. Compared with the original YOLOX-s model, the proposed model takes a two-times higher-resolution input image for higher mean average precision (mAP) using 57% model parameters and 71% computation in Gflops. The proposed model was tested using real drone reconnaissance images, and provided significant improvement in detecting small vehicles.

AI Model for Detection of Abdominal Hemorrhage Lesions in Abdominal CT Images

“…The FPN is separated by the layer specified by the CNN to generate a feature map, and the closer it is to the input layer, the higher the resolution that retains the feature information, such as edges and curves of the image, and the farther it is from the input layer, the less it loses the features that can be inferred through texture or parts of the object. The detection of small objects using DL is a challenging task in computer vision, which requires sufficiently high-resolution images to recognize small objects [ 40 ]. The neural network developed in this study adds a C2 layer to the convolutional network and a P2 layer to the pyramid network to increase the resolution of small lesions by changing the number of utilized layers to four levels between C2 and C5, and six levels between P2 and P7.…”

Soft-NMS-Enabled YOLOv5 with SIOU for Small Water Surface Floater Detection in UAV-Captured Images

“…Kim et al inserted a high-resolution processing module (HRPM) and a sigmoid fusion module (SFM), which not only reduced computational complexity but also improved the detection accuracy of small targets. They obtained good detection results in drone reconnaissance images and small vehicles [40]. Wang et al proposed a bidirectional attention network called BANet, which solved the problems of inaccurate and inefficient detection of small and multiple targets.…”