Learning Rich Features at High-Speed for Single-Shot Object Detection, ICCV, 2019 pytorchpyramidobject-detectionsingle-shot-detection UpdatedDec 10, 2019 Python lars76/chinese-subtitle-ocr Star110 Code Issues Pull requests Optical character recognition for Chinese subtitles using SSD and CNN ...
Single-shot Object Detection 以下转自:http://lanbing510.info/2017/08/28/YOLO-SSD.html 在深度学习出现之前,传统的目标检测方法大概分为区域选择(滑窗)、特征提取(SIFT、HOG等)、分类器(SVM、Adaboost等)三个部分,其主要问题有两方面:一方面滑窗选择策略没有针对性、时间复杂度高,窗口冗余;另一方面手工设计...
[1] Joseph Redmon, Santosh Divvala, Ross Girshick, Ali Farhadi: “You Only Look Once: Unified, Real-Time Object Detection”, 2015;arXiv:1506.02640. [2] Wei Liu, Dragomir Anguelov, Dumitru Erhan, Christian Szegedy, Scott Reed, Cheng-Yang Fu: “SSD: Single Shot MultiBox Detector”, 2016;...
adaptively spatial feature fusion (ASFF)不同尺寸feats的不一致性对检测是个限制。提出ASFF在空间上过滤冲突信息,抑制不一致信息,提高feats的尺度不变性。这个模块的引入在推理上几乎没有开销。 为每个scale的…
论文链接:Feature Selective Anchor-Free Module for Single-Shot Object Detection CVPR2019的一篇single-stage detection的文章,来自CMU。 【Motivation】 目标检测中物体尺度问题一直是个难解决的问题,目前为止主要是从网络结构设计、损失函数、训练方式等方面去缓解尺度带来的烦恼,特别是小物体检测,至今没有一个好的解决...
RefineDet是CVPR 2018的一篇论文,文中提出了一个新的single-shot检测器RefineDet,实现了比二阶段方法更高的准确率而且具有与一阶段方法相当的效率。RefineDet包括两个互连模型ARM(anchor refinement module)和ODM(object detection module):前者用于滤除negative anchors来减少分类器的搜索空间,粗略调整anchors的位置和大小给...
Light Weight Single-Shot Refinement Neural Network for Object Detection. Convolutional neural network based methods have dominated object detection in recent years, which can be divided into the one-stage approach and the two-st... H Desai 被引量: 0发表: 2019年 Faster region based convolution ne...
深度学习论文:Learning Spatial Fusion for Single-Shot Object Detection及其PyTorch实现 https://github.com/shanglianlm0525/PyTorch-Networks 1 概述 本文提出了一种新的数据驱动的自适应空间特征融合(ASFF)金字塔特征融合方式, 通过学习空间上的过滤冲突信息以抑制梯度反传的时候不一致的方法,从而改善了...
Feature Selective Anchor-Free Module for Single-Shot Object Detection 机器学习 提出了一种简单有效的单阶段目标检测模块——特征选择无锚定(FSAF)模块。它可以插入到具有特征金字塔结构的单阶段检测器中。FSAF模块解决了传统基于锚点检测的两个局限性:1)启发式引导的特征选择;2)基于覆盖锚取样。FSAF模块的总体思想...
Object detection has been widely applied in various fields with the rapid development of deep learning in recent years. However, detecting small objects is still a challenging task because of the limited information in features and the complex background