Revisiting RCNN: On Awakening the Classification Power of Faster RCNN

Computer Vision


  • Bowen Cheng
  • Yunchao Wei
  • Honghui Shi
  • Rogerio Feris
  • Jinjun Xiong
  • Thomas Huang

Published on



Computer Vision

Recent region-based object detectors are usually built with separate classification and localization branches on top of shared feature extraction networks. In this paper, we analyze failure cases of state-of-the-art detectors and observe that most hard false positives result from classification instead of localization. We conjecture that: (1) Shared feature representation is not optimal due to the mismatched goals of feature learning for classification and localization; (2) multi-task learning helps, yet optimization of the multi-task loss may result in sub-optimal for individual tasks; (3) large receptive field for different scales leads to redundant context information for small objects.We demonstrate the potential of detector classification power by a simple, effective, and widely-applicable Decoupled Classification Refinement (DCR) network. DCR samples hard false positives from the base classifier in Faster RCNN and trains a RCNN-styled strong classifier. Experiments show new state-of-the-art results on PASCAL VOC and COCO without any bells and whistles.

Please cite our work using the BibTeX below.

  author    = {Bowen Cheng and
               Yunchao Wei and
               Honghui Shi and
               Rog{\'{e}}rio Schmidt Feris and
               Jinjun Xiong and
               Thomas S. Huang},
  title     = {Revisiting {RCNN:} On Awakening the Classification Power of Faster
  journal   = {CoRR},
  volume    = {abs/1803.06799},
  year      = {2018},
  url       = {},
  archivePrefix = {arXiv},
  eprint    = {1803.06799},
  timestamp = {Mon, 13 Aug 2018 16:47:05 +0200},
  biburl    = {},
  bibsource = {dblp computer science bibliography,}
Close Modal