RetinaNet 网络结构（RetinaNet Detector）和实现方式

最新推荐文章于 2024-08-11 09:35:58 发布

Gallant Hu

最新推荐文章于 2024-08-11 09:35:58 发布

阅读量1.2k

点赞数

分类专栏：目标检测从基础到实践--系列论文讲解

本文链接：https://blog.csdn.net/weixin_42108090/article/details/108275666

版权

目标检测从基础到实践--系列论文讲解专栏收录该内容

53 篇文章 14 订阅 ¥59.90 ¥99.00

订阅专栏

超级会员免费看

RetinaNet是一个包含backbone网络和两个特定任务子网的统一网络，用于对象分类和边界框回归。其关键结构是FPN，提供多尺度特征金字塔。每个金字塔层用于检测不同尺度的对象。RetinaNet使用类似RPN的翻译不变锚点，并根据IoU进行分类和边框回归任务的分配。分类子网和边框回归子网各自独立，但共享结构。

摘要由CSDN通过智能技术生成

在这里插入图片描述
RetinaNet is a single, unified network composed of a backbone network and two task-specific subnetworks. The backbone is responsible for computing a convolutional feature map over an entire input image and is an off-the-self convolutional network. The first subnet performs convolutional object classification on the backbone’s output; the second subnet performs convolutional bounding box regression. The two subnetworks feature a simple design that we propose specifically for one-stage, dense detection, see Figure 3. While there are many possible choices for the details of these components, most design parameters are not particularly sensitive to exact values as shown