Weakly Supervised Deep Detection Networks

on Deep Learning

这篇论文被收录在CVPR-2016

Figure 1

图4 Weakly Supervised Deep Detection Network.

摘要

作者提出一个弱监督深度检测结构，在图片分类级别的输入情况下，同时实现区域选择和分类。训练的时候作为图像分类器，隐含的学习目标检测器。

Motivation

pre-trained CNNs在大量的其他任务上泛化性都很好，说明它应该包含一些有意义的数据的表示。CNNs在训练作为图像分类的同时可能已经隐含的包含了执行目标检测的信息。

参考文献

[1] Gong, B., Shi, Y., Sha, F., Grauman, K.: Geodesic flow kernel for unsupervised domain adaptation. In: CVPR (2012)

[2] Gopalan, R., Li, R., Chellappa, R.: Domain adaptation for object recognition: An unsupervised approach. In: ICCV (2011)

[3] Nair, V., Clark, J.J.: An unsupervised, online learning framework for moving object detection. In: CVPR (2004)

[4] Rosenberg, C., Hebert, M., Schneiderman, H.: Semi-supervised self-training of ob- ject detection models. In: WACV (2005)

[5] Wang, X., Hua, G., Han, T.X.: Detection by detections: Non-parametric detector adaptation for a video. In: CVPR (2012)

[6] Wang, M., Wang, X.: Automatic adaptation of a generic pedestrian detector to a specific traffic scene. In: CVPR (2011)

[7] Wang, X., Wang, M., Li, W.: Scene-specific pedestrian detection for static video surveillance. TPAMI 36, 361–374 (2014)

[8] Wu, B., Nevatia, R.: Improving part based object detection by unsupervised, online boosting. In: CVPR (2007)

[9] Levin, A., Viola, P., Freund, Y.: Unsupervised improvement of visual detectors using cotraining. In: ICCV (2003)

☰