Method

Cascade R-CNN: High Quality Object Detection and Instance Segmentation [Cascade MS-CNN]
https://github.com/zhaoweicai/mscnn

Submitted on 2 Jul. 2019 09:23 by
Zhaowei Cai (UCSD)

Running time:0.25 s
Environment:GPU @ 2.5 Ghz (C/C++)

Method Description:
Please check the papers of "Cascade R-CNN: High
Quality Object Detection and Instance
Segmentation", and "A unified multi-scale deep
convolutional neural network for fast object
detection".
Parameters:
Please check the papers.
Latex Bibtex:
@article{cai2019cascade,
title={Cascade R-CNN: High Quality Object
Detection and Instance Segmentation},
author={Cai, Zhaowei and Vasconcelos, Nuno},
journal={arXiv preprint arXiv:1906.09756},
year={2019}
}
@inproceedings{cai2016unified,
title={A unified multi-scale deep
convolutional neural network for fast object
detection},
author={Cai, Zhaowei and Fan, Quanfu and
Feris, Rogerio S and Vasconcelos, Nuno},
booktitle={European conference on computer
vision},
pages={354--370},
year={2016},
organization={Springer}
}

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) 90.68 % 89.95 % 78.40 %
This table as LaTeX


2D object detection results.
This figure as: png eps pdf txt gnuplot




eXTReMe Tracker