English

PFDet: 2nd Place Solution to Open Images Challenge 2018 Object Detection Track

Computer Vision and Pattern Recognition 2018-09-05 v1

Abstract

We present a large-scale object detection system by team PFDet. Our system enables training with huge datasets using 512 GPUs, handles sparsely verified classes, and massive class imbalance. Using our method, we achieved 2nd place in the Google AI Open Images Object Detection Track 2018 on Kaggle.

Keywords

Cite

@article{arxiv.1809.00778,
  title  = {PFDet: 2nd Place Solution to Open Images Challenge 2018 Object Detection Track},
  author = {Takuya Akiba and Tommi Kerola and Yusuke Niitani and Toru Ogawa and Shotaro Sano and Shuji Suzuki},
  journal= {arXiv preprint arXiv:1809.00778},
  year   = {2018}
}

Comments

Technical report for Open Images Challenge 2018 Object Detection Track

R2 v1 2026-06-23T03:53:14.354Z