LWANet

Attention-Guided Lightweight Network for Real-Time Segmentation of Robotic Surgical Instruments (ICRA2020)

LWANet can segment surgical instruments in real-time while takes little computational costs. Based on 960×544 inputs, its inference speed can reach 39 fps with only 3.39 GFLOPs. Also, it has a small model size and the number of parameters is only 2.06 M. The proposed network is evaluated on two datasets. It achieves state-of-the-art performance 94.10% mean IOU on Cata7 and obtains a new record on EndoVis 2017 with a 4.10% increase on mean IOU.

Results

Cata7
table1

table2
EndoVis 2017
table3

Citation

If you find LWANet useful in your research, please consider citing:

@article{ni2019attention,
  title={Attention-guided lightweight network for real-time segmentation of robotic surgical instruments},
  author={Ni, Zhen-Liang and Bian, Gui-Bin and Hou, Zeng-Guang and Zhou, Xiao-Hu and Xie, Xiao-Liang and Li, Zhen},
  journal={arXiv preprint arXiv:1910.11109},
  year={2019}
}

GitHub