Paper_CV_3 目标检测

[toc]

计算机视觉任务：

图像分类、目标定位、目标检测、目标跟踪，语义分割，实例分割；

detection

目标检测领域中主流的两大类方法:

dense detector: 例如DPM，YOLO，RetinaNet，FCOS。在dense detector中，大量的object candidates例如sliding-windows，anchor-boxes， reference-points等被提前预设在图像网格或者特征图网格上，然后直接预测这些candidates到gt的scaling/offest和物体类别。

dense-to-sparse detector: RCNN家族，对一组sparse的candidates预测和分类

Detection模型算法整理

RCNN内容整理RCNN, FastRCNN,FasterRCNN,YOLO,SSD

发展历史：

R-CNN

Fast R-CNN

Faster R-CNN

Mask R-CNN （目标检测 + 像素分割）

SSD (Single Shot MultiBox Defender)

YOLO (You Only Look Once)

YOLO_V2

YOLO_V3

DSSD

IoU-Net(旷视科技)

SPPNet: Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition, ECCV, 2014
 R-CNN: Rich feature hierarchies for accurate object detection and semantic segmentation, CVPR, 2014
 Fast R-CNN, ICCV, 2015
 Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks, NIPS, 2015
 FPN: Feature Pyramid Networks for Object Detection, CVPR, 2017
 SSD: Single Shot MultiBox Detector, ECCV, 2016
 DSSD: Deconvolutional Single Shot Detector, CVPR, 2017
 FSSD: Feature Fusion Single Shot Multibox Detector, arXiv, 2017
 YOLO-v1: You Only Look Once: Unified, Real-Time Object Detection, ECCV, 2016
 YOLO-v2: YOLO9000: Better, Faster, Stronger, arXiv, 2016
 YOLO-v3: YOLOv3: An Incremental Improvement
 R-FCN: Object Detection via Region-based Fully Convolutional Networks, NIPS, 2016
 Deformable Convolutional Networks, ICCV, 2017
 Faster R-CNN+SSD: Single-Shot Refinement Neural Network for Object Detection, CVPR, 2018
 CornerNet: Detecting Objects as Paired Keypoints, ECCV, 2018
 IoUNet: Acquisition of Localization Confidence for Accurate Object Detection, ECCV, 2018