归档: 2018/5

源码分析-TensorFlow目标检测API

参考: [1]TensorFlow Object Detection API: basics of detection: https://becominghuman.ai/tensorflow-object-detection-api-basics-of-detection-7b134d689c75 https://becominghuman.ai/tensorflow-object-detect

论文-anchor设计

S3FD:Single Shot Scale-invariant Face Detector参考:https://zhuanlan.zhihu.com/p/29022585; 微信2的收藏; 存在的问题: 在detection layer上,small face最后拥有的特征太少; anchor与RF的尺度不匹配; tiny face和outer face不能匹配到足够多的anchors; sma

目标检测概述(待完善)

参考: https://zhuanlan.zhihu.com/p/33544892 https://kuaibao.qq.com/s/20180319G1QL8500?refer=spider 目标检测近年来已经取得了很重要的进展,主流的算法主要分为两个类型(参考RefineDet): (1)two-stage方法,如R-CNN系算法,其主要思路是先通过启发式方法(selective searc

mAP评估指标

参考: https://www.zhihu.com/question/53405779/answer/419532990 概念及公式11点插值法计算 实例:计算mAPhttps://github.com/rafaelpadilla/Object-Detection-Metrics coco目标检测评估标准https://blog.csdn.net/u014734886/article/de