目标检测之YOLOV 3论文阅读笔记

最新推荐文章于 2023-01-13 08:00:00 发布

专注于计算机视觉的AndyJiang

最新推荐文章于 2023-01-13 08:00:00 发布

阅读量147

点赞数

分类专栏：计算机视觉文章标签：深度学习计算机视觉

本文链接：https://blog.csdn.net/andyjkt/article/details/107387567

版权

计算机视觉专栏收录该内容

31 篇文章 10 订阅

订阅专栏

提出时间：2018年，a tech report
网络结构如下
backbone－Feature Extractor
darknet53
在这里插入图片描述
Bounding Box Prediction

Class Prediction
multilabel classification use independent logistic classifiers

Predictions Across Scales
每一个尺度都有，共三个尺度，每一个尺度3个bounding boxes．
N × N × [3 ∗ (4 + 1 + 80)] for the 4 bounding box offsets, 1 objectness prediction, and 80 class predictions.

聚类中心为：
(10 × 13), (16 × 30), (33 × 23), (30 × 61), (62 × 45), (59 × 119), (116 × 90), (156 × 198), (373 × 326).

一些不work的实验：
Anchor box x, y offset predictions.
Linear x, y predictions instead of logistic
Focal loss.
Dual IOU thresholds and truth assignment.
Faster R-CNN uses two IOU thresholds during training. If a prediction overlaps the ground truth by .7 it is as a positive example, by [.3 − .7] it is ignored, less than .3 for all ground truth objects it is a negative example. We tried a similar strategy but couldn’t get good results.
在这里插入图片描述
results

AP指标和SSD差不多，AP0.5和RetinaNet相当，AP0.75性能下降，APs(小尺度)好于SSD,中大尺度没有优势，之前的yolo在小尺度上是劣势