Skip to content

Latest commit

 

History

History
 
 

nas_fpn

NAS-FPN: Learning Scalable Feature Pyramid Architecture for Object Detection

Abstract

Current state-of-the-art convolutional architectures for object detection are manually designed. Here we aim to learn a better architecture of feature pyramid network for object detection. We adopt Neural Architecture Search and discover a new feature pyramid architecture in a novel scalable search space covering all cross-scale connections. The discovered architecture, named NAS-FPN, consists of a combination of top-down and bottom-up connections to fuse features across scales. NAS-FPN, combined with various backbone models in the RetinaNet framework, achieves better accuracy and latency tradeoff compared to state-of-the-art object detection models. NAS-FPN improves mobile detection accuracy by 2 AP compared to state-of-the-art SSDLite with MobileNetV2 model in [32] and achieves 48.3 AP which surpasses Mask R-CNN [10] detection accuracy with less computation time.

Citation

@inproceedings{ghiasi2019fpn,
  title={Nas-fpn: Learning scalable feature pyramid architecture for object detection},
  author={Ghiasi, Golnaz and Lin, Tsung-Yi and Le, Quoc V},
  booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
  pages={7036--7045},
  year={2019}
}

Results and Models

We benchmark the new training schedule (crop training, large batch, unfrozen BN, 50 epochs) introduced in NAS-FPN. RetinaNet is used in the paper.

Backbone Lr schd Mem (GB) Inf time (fps) box AP Config Download
R-50-FPN 50e 12.9 22.9 37.9 config model | log
R-50-NASFPN 50e 13.2 23.0 40.5 config model | log

Note: We find that it is unstable to train NAS-FPN and there is a small chance that results can be 3% mAP lower.