Method Details
Details for method 'Axial-DeepLab-L [Cityscapes-fine]'
Method overview
| name | Axial-DeepLab-L [Cityscapes-fine] |
| challenge | instance-level semantic labeling |
| details | Convolution exploits locality for efficiency at a cost of missing long range context. Self-attention has been adopted to augment CNNs with non-local interactions. Recent works prove it possible to stack self-attention layers to obtain a fully attentional network by restricting the attention to a local region. In this paper, we attempt to remove this constraint by factorizing 2D self-attention into two 1D self-attentions. This reduces computation complexity and allows performing attention within a larger or even global region. In companion, we also propose a position-sensitive self-attention design. Combining both yields our position-sensitive axial-attention layer, a novel building block that one could stack to form axial-attention models for image classification and dense prediction. We demonstrate the effectiveness of our model on four large-scale datasets. In particular, our model outperforms all existing stand-alone self-attention models on ImageNet. Our Axial-DeepLab improves 2.8% PQ over bottom-up state-of-the-art on COCO test-dev. This previous state-of-the-art is attained by our small variant that is 3.8x parameter-efficient and 27x computation-efficient. Axial-DeepLab also achieves state-of-the-art results on Mapillary Vistas and Cityscapes. |
| publication | Axial-DeepLab: Stand-Alone Axial-Attention for Panoptic Segmentation Huiyu Wang, Yukun Zhu, Bradley Green, Hartwig Adam, Alan Yuille, Liang-Chieh Chen ECCV 2020 (spotlight) https://arxiv.org/abs/2003.07853 |
| project page / code | https://github.com/csrhddlam/axial-deeplab |
| used Cityscapes data | fine annotations |
| used external data | ImageNet |
| runtime | n/a |
| subsampling | no |
| submission date | March, 2020 |
| previous submissions |
Average results
| Metric | Value |
|---|---|
| AP | 33.2517 |
| AP50% | 54.9438 |
| AP100m | 48.7572 |
| AP50m | 51.9623 |
Class results
| Class | AP | AP50% | AP100m | AP50m |
|---|---|---|---|---|
| person | 31.9704 | 60.4837 | 51.2816 | 51.3372 |
| rider | 27.8489 | 60.5161 | 42.3996 | 42.9868 |
| car | 52.3964 | 74.0389 | 74.4562 | 77.148 |
| truck | 30.6836 | 39.4906 | 45.1815 | 54.3969 |
| bus | 44.1933 | 55.9174 | 66.5918 | 76.6773 |
| train | 33.966 | 53.4695 | 45.2847 | 47.3518 |
| motorcycle | 25.7373 | 50.1678 | 34.8293 | 35.6792 |
| bicycle | 19.2181 | 45.4667 | 30.0332 | 30.1213 |