Dual-Erp Representation for Object Detection in 360° Images

Miao Cao, Satoshi Ikehata, Kiyoharu AizawaICIP 2022 · OpenReview

Abstract

Object detection has achieved good performance on perspective images. However, a general object detector does not maintain this performance when applied to a 360° image in a single equirectangular projection (ERP) or multi-projection representation because of the distortion in the high-latitude region or discontinuity at the boundaries. In this paper, we proposed dual-ERP, which is a multi-view ERP representation, as the network input for 360° object detection in training and inference. Dual-ERP combines the advantages of single ERP and multi-projection representations, and it can easily be integrated with existing object detectors. The experimental results showed that compared to other representations, dual-ERP significantly improved the performance of different baseline object detectors.
BibTeX
@inproceedings{cao2022dual,
  author = {Miao Cao and Satoshi Ikehata and Kiyoharu Aizawa},
  title = {{Dual-Erp Representation for Object Detection in 360° Images}},
  booktitle = {ICIP},
  year = {2022},
  url = {https://openreview.net/forum?id=guRTPFlGHfV},
}

Loading related content...

Figure