FishNet

This repo holds the implementation code of the paper:

FishNet: A Versatile Backbone for Image, Region, and Pixel Level Prediction , Shuyang Sun, Jiangmiao Pang, Jianping Shi, Shuai Yi, Wanli Ouyang, NeurIPS 2018.

FishNet was used as a key component for winning the 1st place in COCO Detection Challenge 2018.

Prerequisites

Python 3.6.x
PyTorch 0.4.0 (0.3.1 support soon)

Data Augmentation

Method	Settings
Random Flip	True
Random Crop	8% ~ 100%
Aspect Ratio	3/4 ~ 4/3
Random PCA Lighting	0.1

Note: We apply weight decay to all weights and biases instead of just the weights of the convolution layers.

Training

To train FishNet-150 with 8 GPUs and batch size 256, simply run

python main.py --config "cfgs/fishnet150.yaml" IMAGENET_ROOT_PATH

Models

Models trained without tricks

Model	Top-1	Top-5	Baidu Yun	Google Cloud
FishNet99	77.41%	93.59%	Download	Download
FishNet150	78.14%	93.95%	Download	Download
FishNet201	78.76%	94.39%	Available Soon	Available Soon

Models trained with cosine lr schedule (200 epochs) and label smoothing

Model	Top-1	Top-5	Baidu Yun	Google Cloud
FishNet99	-	-	Available Soon	Available Soon
FishNet150	79.35%	94.75%	Download	Download
FishNet201	-	-	Available Soon	Available Soon

TODO:

Update our arxiv paper.
Release pre-train models.
Train the model with more training tricks.

Citation

If you find our research useful, please cite the paper:

@inproceedings{sun2018fishnet,
  title={FishNet: A Versatile Backbone for Image, Region, and Pixel Level Prediction},
  author={Sun, Shuyang and Pang, Jiangmiao and Shi, Jianping and Yi, Shuai and Ouyang, Wanli},
  booktitle={Advances in Neural Information Processing Systems},
  pages={760--770},
  year={2018}
}

Contact

You can contact Shuyang Sun by sending email to shuyang.sun@sydney.edu.au.

ruc98/FishNet