目标检测,也叫目标提取,是一种基于目标几何和统计特征的图像分割,它将目标的分割和识别合二为一,其准确性和实时性是整个系统的一项重要能力。尤其是在复杂场景中,需要对多个目标进行实时处理时,目标自动提取和识别就显得特别重要。 随着计算机技术的发展和计算机视觉原理的广泛应用,利用计算机图像处理技术对目标进行实时跟踪研究越来越热门,对目标进行动态实时跟踪定位在智能化交通系统、智能监控系统、军事目标检测及医学导航手术中手术器械定位等方面具有广泛的应用价值。

目标检测(物体检测, Object Detection) 专知荟萃

入门学习

  1. 图像目标检测(Object Detection)原理与实现 (1-6)

  2. . 基于特征共享的高效物体检测 Faster R-CNN和ResNet的作者任少卿 博士毕业论文 中文

  3. R-CNN:论文笔记

  4. Fast-RCNN:

  5. Faster-RCNN:

  6. FPN:

  7. R-FCN:

  8. SSD:

  9. YOLO:

  10. DenseBox:余凯特邀报告:基于密集预测图的物体检测技术造就全球领先的ADAS系统

  11. PVANET: Deep but Lightweight Neural Networks for Real-time Object Detection - [http://www.cnblogs.com/xueyuxiaolang/p/5959442.html]

  12. 深度学习论文笔记:DSSD - [http://jacobkong.github.io/posts/2938514597/]

  13. DSOD

  14. Focal Loss:

  15. Soft-NMS:

  16. OHEM:

  17. Mask-RCNN 2017:

  18. 目标检测之比较

  19. 视觉目标检测和识别之过去,现在及可能

进阶文章

  1. Deep Neural Networks for Object Detection (基于DNN的对象检测)NIPS2013:
  2. R-CNN Rich feature hierarchies for accurate object detection and semantic segmentation:
  3. Fast R-CNN :
  4. Faster R-CNN Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks:
  5. Scalable Object Detection using Deep Neural Networks
  6. Scalable, High-Quality Object Detection
  7. SPP-Net Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition
  8. DeepID-Net DeepID-Net: Deformable Deep Convolutional Neural Networks for Object Detection
  9. Object Detectors Emerge in Deep Scene CNNs
  10. segDeepM: Exploiting Segmentation and Context in Deep Neural Networks for Object Detection
  11. Object Detection Networks on Convolutional Feature Maps
  12. Improving Object Detection with Deep Convolutional Networks via Bayesian Optimization and Structured Prediction
  13. DeepBox: Learning Objectness with Convolutional Networks
  14. Object detection via a multi-region & semantic segmentation-aware CNN model
  15. You Only Look Once: Unified, Real-Time Object Detection
  16. YOLOv2 YOLO9000: Better, Faster, Stronger
  17. AttentionNet: Aggregating Weak Directions for Accurate Object Detection
  18. DenseBox: Unifying Landmark Localization with End to End Object Detection
  19. SSD: Single Shot MultiBox Detector
  20. DSSD : Deconvolutional Single Shot Detector
  21. G-CNN: an Iterative Grid Based Object Detector
  22. HyperNet: Towards Accurate Region Proposal Generation and Joint Object Detection
  23. A MultiPath Network for Object Detection
  24. R-FCN: Object Detection via Region-based Fully Convolutional Networks
  25. A Unified Multi-scale Deep Convolutional Neural Network for Fast Object Detection
  26. PVANET: Deep but Lightweight Neural Networks for Real-time Object Detection
  27. Feature Pyramid Networks for Object Detection
  28. Learning Chained Deep Features and Classifiers for Cascade in Object Detection
  29. DSOD: Learning Deeply Supervised Object Detectors from Scratch
  30. Focal Loss for Dense Object Detection  ICCV 2017 Best student paper award. Facebook AI Research

综述

  1. 深度学习之 "物体检测" 方法梳理

  2. 地平线黄李超开讲:深度学习和物体检测!:

  3. 对话CVPR2016:目标检测新进展:

  4. 基于深度学习的目标检测技术演进:R-CNN、Fast R-CNN、Faster R-CNN:

  5. 基于深度学习的目标检测研究进展

  6. 讲堂干货No.1|山世光-基于深度学习的目标检测技术进展与展望

Tutorial

  1. CVPR'17 Tutorial Deep Learning for Objects and Scenes by Kaiming He Ross Girshick
  2. ICCV 2015 Tools for Efficient Object Detection
  3. Object Detection
  4. Image Recognition and Object Detection : Part 1
  5. R-CNN for Object Detection

视频教程

  1. cs231 第11讲 Detection and Segmentation 
  2. Deep Learning for Instance-level Object Understanding by Ross Girshick.

代码

  1. R-CNN
  2. Fast R-CNN:
  3. Faster R-CNN
  4. SPP-Net
  5. YOLO
  6. YOLOv2
  7. SSD
  8. Recurrent Scale Approximation for Object Detection in CNN

  9. Mask-RCNN 2017

领域专家

  1. Ross Girshick (rbg 大神)
  2. Kaiming He, Facebook人工智能实验室科学家Kaiming He
  3. Shaoqing Ren
  4. Jian Sun

初步版本,水平有限,有错误或者不完善的地方,欢迎大家提建议和补充,会一直保持更新,本文为专知内容组原创内容,未经允许不得转载,如需转载请发送邮件至[email protected] 或 联系微信专知小助手(Rancho_Fang)

敬请关注http://www.zhuanzhi.ai 和关注专知公众号,获取第一手AI相关知识

展开全文
Top