Bridging the Gap Between Anchor-based and Anchor-free Detection via Adaptive Training Sample Selection
Shifeng Zhang1,2, Cheng Chi3, Yongqiang Yao4, Zhen Lei1,2, Stan Z. Li5 1 CBSR, NLPR, CASIA 2 SAI, UCAS 3 AIR, CAS 4 BUPT 5 Westlake University Corresponding author
Abstract
Object detection has been dominated by anchor-based detectors for several years. Recently, anchor-free detectors have become popular due to the proposal of FPN and Focal Loss. In this paper, we first point out that the essential difference between anchor-based and anchor-free detection is actually how to define positive and negative training samples, which leads to the performance gap between them. If they adopt the same definition of positive and negative samples during training, there is no obvious difference in the final performance, no matter regressing from a box or a point. This shows that how to select positive and negative training samples is important for current object detectors. Then, we propose an Adaptive Training Sample Selection (ATSS) to automatically select positive and negative samples according to statistical characteristics of object. It significantly improves the performance of anchor-based and anchor-free detectors and bridges the gap between them. Finally, we discuss the necessity of tiling multiple anchors per location on the image to detect objects. Extensive experiments conducted on MS COCO support our aforementioned analysis and conclusions. With the newl
中文速览
目标检测领域长期被"基于锚框（anchor-based）"的方法统治，而近年兴起的"无锚（anchor-free）"方法虽然性能相近却机制不同，两者差距的根源一直不明。作者通过严格的对照实验发现，两类方法的核心区别其实只有一点：如何定义训练时的正负样本，而非锚框的有无或回归起点的差异——一旦统一正负样本的划分规则，两者性能几乎持平。基于这一洞察，作者提出了自适应训练样本选择（Adaptive Training Sample Selection，ATSS）方法，根据每个目标的统计特性自动确定IoU阈值来筛选正负样本，既不需要手工调参，也不引入任何额外计算开销。在MS COCO基准上，引入ATSS后检测精度大幅提升至50.7% AP，刷新了当时的最优水平；这项工作不仅统一了两类检测范式的理论认知，更为后续检测器设计提供了清晰的改进方向。
原文 arXiv:1912.02424;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1912.02424v4