Spaces:
Runtime error
Runtime error
File size: 14,151 Bytes
412c852 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 |
# ANN
> [Asymmetric Non-local Neural Networks for Semantic Segmentation](https://arxiv.org/abs/1908.07678)
## Introduction
<!-- [ALGORITHM] -->
<a href="https://github.com/MendelXu/ANN">Official Repo</a>
<a href="https://github.com/open-mmlab/mmsegmentation/blob/v0.17.0/mmseg/models/decode_heads/ann_head.py#L185">Code Snippet</a>
## Abstract
<!-- [ABSTRACT] -->
The non-local module works as a particularly useful technique for semantic segmentation while criticized for its prohibitive computation and GPU memory occupation. In this paper, we present Asymmetric Non-local Neural Network to semantic segmentation, which has two prominent components: Asymmetric Pyramid Non-local Block (APNB) and Asymmetric Fusion Non-local Block (AFNB). APNB leverages a pyramid sampling module into the non-local block to largely reduce the computation and memory consumption without sacrificing the performance. AFNB is adapted from APNB to fuse the features of different levels under a sufficient consideration of long range dependencies and thus considerably improves the performance. Extensive experiments on semantic segmentation benchmarks demonstrate the effectiveness and efficiency of our work. In particular, we report the state-of-the-art performance of 81.3 mIoU on the Cityscapes test set. For a 256x128 input, APNB is around 6 times faster than a non-local block on GPU while 28 times smaller in GPU running memory occupation. Code is available at: [this https URL](https://github.com/MendelXu/ANN).
<!-- [IMAGE] -->
<div align=center>
<img src="https://user-images.githubusercontent.com/24582831/142898322-3bbd578c-e488-4bae-9c14-7598adac5cbd.png" width="70%"/>
</div>
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | Device | mIoU | mIoU(ms+flip) | config | download |
| ------ | -------- | --------- | ------: | -------- | -------------- | ------ | ----: | ------------: | ------------------------------------------------------------------------------------------------------------------------ | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| ANN | R-50-D8 | 512x1024 | 40000 | 6 | 3.71 | V100 | 77.40 | 78.57 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/ann/ann_r50-d8_4xb2-40k_cityscapes-512x1024.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_512x1024_40k_cityscapes/ann_r50-d8_512x1024_40k_cityscapes_20200605_095211-049fc292.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_512x1024_40k_cityscapes/ann_r50-d8_512x1024_40k_cityscapes_20200605_095211.log.json) |
| ANN | R-101-D8 | 512x1024 | 40000 | 9.5 | 2.55 | V100 | 76.55 | 78.85 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/ann/ann_r101-d8_4xb2-40k_cityscapes-512x1024.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_512x1024_40k_cityscapes/ann_r101-d8_512x1024_40k_cityscapes_20200605_095243-adf6eece.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_512x1024_40k_cityscapes/ann_r101-d8_512x1024_40k_cityscapes_20200605_095243.log.json) |
| ANN | R-50-D8 | 769x769 | 40000 | 6.8 | 1.70 | V100 | 78.89 | 80.46 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/ann/ann_r50-d8_4xb2-40k_cityscapes-769x769.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_769x769_40k_cityscapes/ann_r50-d8_769x769_40k_cityscapes_20200530_025712-2b46b04d.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_769x769_40k_cityscapes/ann_r50-d8_769x769_40k_cityscapes_20200530_025712.log.json) |
| ANN | R-101-D8 | 769x769 | 40000 | 10.7 | 1.15 | V100 | 79.32 | 80.94 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/ann/ann_r101-d8_4xb2-40k_cityscapes-769x769.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_769x769_40k_cityscapes/ann_r101-d8_769x769_40k_cityscapes_20200530_025720-059bff28.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_769x769_40k_cityscapes/ann_r101-d8_769x769_40k_cityscapes_20200530_025720.log.json) |
| ANN | R-50-D8 | 512x1024 | 80000 | - | - | V100 | 77.34 | 78.65 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/ann/ann_r50-d8_4xb2-80k_cityscapes-512x1024.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_512x1024_80k_cityscapes/ann_r50-d8_512x1024_80k_cityscapes_20200607_101911-5a9ad545.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_512x1024_80k_cityscapes/ann_r50-d8_512x1024_80k_cityscapes_20200607_101911.log.json) |
| ANN | R-101-D8 | 512x1024 | 80000 | - | - | V100 | 77.14 | 78.81 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/ann/ann_r101-d8_4xb2-80k_cityscapes-512x1024.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_512x1024_80k_cityscapes/ann_r101-d8_512x1024_80k_cityscapes_20200607_013728-aceccc6e.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_512x1024_80k_cityscapes/ann_r101-d8_512x1024_80k_cityscapes_20200607_013728.log.json) |
| ANN | R-50-D8 | 769x769 | 80000 | - | - | V100 | 78.88 | 80.57 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/ann/ann_r50-d8_4xb2-80k_cityscapes-769x769.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_769x769_80k_cityscapes/ann_r50-d8_769x769_80k_cityscapes_20200607_044426-cc7ff323.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_769x769_80k_cityscapes/ann_r50-d8_769x769_80k_cityscapes_20200607_044426.log.json) |
| ANN | R-101-D8 | 769x769 | 80000 | - | - | V100 | 78.80 | 80.34 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/ann/ann_r101-d8_4xb2-80k_cityscapes-769x769.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_769x769_80k_cityscapes/ann_r101-d8_769x769_80k_cityscapes_20200607_013713-a9d4be8d.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_769x769_80k_cityscapes/ann_r101-d8_769x769_80k_cityscapes_20200607_013713.log.json) |
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | Device | mIoU | mIoU(ms+flip) | config | download |
| ------ | -------- | --------- | ------: | -------- | -------------- | ------ | ----: | ------------: | -------------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| ANN | R-50-D8 | 512x512 | 80000 | 9.1 | 21.01 | V100 | 41.01 | 42.30 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/ann/ann_r50-d8_4xb4-80k_ade20k-512x512.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_512x512_80k_ade20k/ann_r50-d8_512x512_80k_ade20k_20200615_014818-26f75e11.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_512x512_80k_ade20k/ann_r50-d8_512x512_80k_ade20k_20200615_014818.log.json) |
| ANN | R-101-D8 | 512x512 | 80000 | 12.5 | 14.12 | V100 | 42.94 | 44.18 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/ann/ann_r101-d8_4xb4-80k_ade20k-512x512.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_512x512_80k_ade20k/ann_r101-d8_512x512_80k_ade20k_20200615_014818-c0153543.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_512x512_80k_ade20k/ann_r101-d8_512x512_80k_ade20k_20200615_014818.log.json) |
| ANN | R-50-D8 | 512x512 | 160000 | - | - | V100 | 41.74 | 42.62 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/ann/ann_r50-d8_4xb4-160k_ade20k-512x512.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_512x512_160k_ade20k/ann_r50-d8_512x512_160k_ade20k_20200615_231733-892247bc.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_512x512_160k_ade20k/ann_r50-d8_512x512_160k_ade20k_20200615_231733.log.json) |
| ANN | R-101-D8 | 512x512 | 160000 | - | - | V100 | 42.94 | 44.06 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/ann/ann_r101-d8_4xb4-160k_ade20k-512x512.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_512x512_160k_ade20k/ann_r101-d8_512x512_160k_ade20k_20200615_231733-955eb1ec.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_512x512_160k_ade20k/ann_r101-d8_512x512_160k_ade20k_20200615_231733.log.json) |
### Pascal VOC 2012 + Aug
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | Device | mIoU | mIoU(ms+flip) | config | download |
| ------ | -------- | --------- | ------: | -------- | -------------- | ------ | ----: | ------------: | --------------------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| ANN | R-50-D8 | 512x512 | 20000 | 6 | 20.92 | V100 | 74.86 | 76.13 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/ann/ann_r50-d8_4xb4-20k_voc12aug-512x512.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_512x512_20k_voc12aug/ann_r50-d8_512x512_20k_voc12aug_20200617_222246-dfcb1c62.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_512x512_20k_voc12aug/ann_r50-d8_512x512_20k_voc12aug_20200617_222246.log.json) |
| ANN | R-101-D8 | 512x512 | 20000 | 9.5 | 13.94 | V100 | 77.47 | 78.70 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/ann/ann_r101-d8_4xb4-20k_voc12aug-512x512.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_512x512_20k_voc12aug/ann_r101-d8_512x512_20k_voc12aug_20200617_222246-2fad0042.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_512x512_20k_voc12aug/ann_r101-d8_512x512_20k_voc12aug_20200617_222246.log.json) |
| ANN | R-50-D8 | 512x512 | 40000 | - | - | V100 | 76.56 | 77.51 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/ann/ann_r50-d8_4xb4-40k_voc12aug-512x512.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_512x512_40k_voc12aug/ann_r50-d8_512x512_40k_voc12aug_20200613_231314-b5dac322.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_512x512_40k_voc12aug/ann_r50-d8_512x512_40k_voc12aug_20200613_231314.log.json) |
| ANN | R-101-D8 | 512x512 | 40000 | - | - | V100 | 76.70 | 78.06 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/ann/ann_r101-d8_4xb4-40k_voc12aug-512x512.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_512x512_40k_voc12aug/ann_r101-d8_512x512_40k_voc12aug_20200613_231314-bd205bbe.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_512x512_40k_voc12aug/ann_r101-d8_512x512_40k_voc12aug_20200613_231314.log.json) |
## Citation
```bibtex
@inproceedings{zhu2019asymmetric,
title={Asymmetric non-local neural networks for semantic segmentation},
author={Zhu, Zhen and Xu, Mengde and Bai, Song and Huang, Tengteng and Bai, Xiang},
booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision},
pages={593--602},
year={2019}
}
```
|