Scale Dependent Pooling with RPN [SDP+RPN]

Submitted on 1 Apr. 2016 21:41 by
Wongun Choi (NEC Laboratories)

Running time:0.4 s
Environment:GPU @ 2.5 Ghz (Python + C/C++)

Method Description:
We applied SDP with RPN (FasterRCNN). Please note that
this result is not included in the paper, as the extension was
done after submission.

Some details:
1. VGG16 architecture for conv layers. The parameters are
initialized by an image-net trained model.
2. 3 RPNs at conv3, conv4, and conv5
3. 3 SDPs at conv3, conv4, and conv5
Latex Bibtex:
title={Exploit All the Layers: Fast and Accurate CNN
Object Detector with Scale Dependent Pooling and
Cascaded Rejection Classifiers},
author={Fan Yang and Wongun Choi and Yuanqing Lin},
booktitle={Proceedings of the IEEE International
Conference on Computer Vision and Pattern
title={Faster R-CNN: Towards real-time object detection
with region proposal networks},
author={Ren, Shaoqing and He, Kaiming and Girshick,
Ross and Sun, Jian},
booktitle={Advances in Neural Information Processing

Detailed Results

Object detection and orientation estimation results. Results for object detection are given in terms of average precision (AP) and results for joint object detection and orientation estimation are provided in terms of average orientation similarity (AOS).

Benchmark Easy Moderate Hard
Car (Detection) 89.90 % 89.42 % 78.54 %
Pedestrian (Detection) 79.98 % 70.20 % 64.84 %
Cyclist (Detection) 81.05 % 73.08 % 64.88 %
This table as LaTeX

This figure as: png eps pdf txt gnuplot

This figure as: png eps pdf txt gnuplot

This figure as: png eps pdf txt gnuplot

eXTReMe Tracker