AUTHOR=Huang Li , Chen Cheng , Yun Juntong , Sun Ying , Tian Jinrong , Hao Zhiqiang , Yu Hui , Ma Hongjie 

TITLE=Multi-Scale Feature Fusion Convolutional Neural Network for Indoor Small Target Detection

JOURNAL=Frontiers in Neurorobotics

VOLUME=Volume 16 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/neurorobotics/articles/10.3389/fnbot.2022.881021

DOI=10.3389/fnbot.2022.881021

ISSN=1662-5218

ABSTRACT=The development of object detection technology makes it possible for robots to interact with people and the environment, but the changeable application scenarios make the detection accuracy of small and medium objects in the practical application of object detection technology is low and the effect is not good. In this paper, based on multi-scale feature fusion of indoor small target detection method, using the device to collect different Angle, light and shade condition indoor image, and use the image enhancement technology for amplification of the data set, set up a data set, indoor scenarios and the SSD algorithm in target detection layer and its adjacent features fusion, in the end, The Faster R-CNN, YOLOv3, SSD and SSD target detection models based on multi-scale feature fusion were trained on indoor scene data set based on transfer learning. The experimental results show that multi-scale feature fusion can improve the detection accuracy of all kinds of objects, especially for objects with relatively small scale. In addition, although the detection speed of the improved SSD algorithm decreases, it is Faster than the Faster R-CNN, which better achieves the balance between target detection accuracy and speed.