Preprint link: ConvMLP: Hierarchical Convolutional MLPs for Vision
Method | Backbone | Lr schd | box AP | mask AP | Config | Download |
---|---|---|---|---|---|---|
RetinaNet | ConvMLP-S | 1x | 37.2 | - | config | model |
RetinaNet | ConvMLP-M | 1x | 39.4 | - | config | model |
RetinaNet | ConvMLP-L | 1x | 40.2 | - | config | model |
Mask R-CNN | ConvMLP-S | 1x | 38.4 | 35.7 | config | model |
Mask R-CNN | ConvMLP-M | 1x | 40.6 | 37.2 | config | model |
Mask R-CNN | ConvMLP-L | 1x | 41.7 | 38.2 | config | model |
@article{li2021convmlp,
title={ConvMLP: Hierarchical Convolutional MLPs for Vision},
author={Jiachen Li and Ali Hassani and Steven Walton and Humphrey Shi},
year={2021},
eprint={2109.04454},
archivePrefix={arXiv},
primaryClass={cs.CV}
}