YOLO-CORE: Contour Regression for Efficient Instance Segmentation

作     者：Haoliang Liu Wei Xiong Yu Zhang Haoliang Liu;Wei Xiong;Yu Zhang

作者机构：School of Computer Science and Engineering and the Key Laboratory of Computer Network and Information Integration(Ministry of Education)Southeast UniversityNanjing 211189China 

出 版 物：《Machine Intelligence Research》 (机器智能研究(英文版))

年 卷 期：2023年第20卷第5期

页      面：716-728页

核心收录：

学科分类：0831[工学-生物医学工程（可授工学、理学、医学学位）] 08[工学] 080203[工学-机械设计及理论] 0802[工学-机械工程] 0714[理学-统计学（可授理学、经济学学位）] 0701[理学-数学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：supported by the National Key R&D Program of China(Nos.2018AAA0100104 and 2018AAA0100100) Natural Science Foundation of Jiangsu Province,China(No.BK20211164). 

主　　题：Computer vision instance segmentation object shape prediction contour regression polar distance. 

摘      要：Instance segmentation has drawn mounting attention due to its significant utility.However,high computational costs have been widely acknowledged in this domain,as the instance mask is generally achieved by pixel-level labeling.In this paper,we present a conceptually efficient contour regression network based on the you only look once(YOLO)architecture named YOLO-CORE for instance segmentation.The mask of the instance is efficiently acquired by explicit and direct contour regression using our designed multiorder constraint consisting of a polar distance loss and a sector loss.Our proposed YOLO-CORE yields impressive segmentation performance in terms of both accuracy and speed.It achieves 57.9%AP@0.5 with 47 FPS(frames per second)on the semantic boundaries dataset(SBD)and 51.1%AP@0.5 with 46 FPS on the COCO dataset.The superior performance achieved by our method with explicit contour regression suggests a new technique line in the YOLO-based image understanding field.Moreover,our instance segmentation design can be flexibly integrated into existing deep detectors with negligible computation cost(65.86 BFLOPs(billion float operations per second)to 66.15 BFLOPs with the YOLOv3 detector).