Method Details
Details for method 'Unifying Training and Inference for Panoptic Segmentation [Cityscapes-fine]'
Method overview
| name | Unifying Training and Inference for Panoptic Segmentation [Cityscapes-fine] |
| challenge | panoptic semantic labeling |
| details | We present an end-to-end network to bridge the gap between training and inference pipeline for panoptic segmentation. In contrast to recent works, our network exploits a parametrised, yet lightweight panoptic segmentation submodule, powered by an end-to-end learnt dense instance affinity, to capture the probability that any pair of pixels belong to the same instance. This panoptic submodule gives rise to a novel propagation mechanism for panoptic logits and enables the network to output a coherent panoptic segmentation map for both “stuff” and “thing” classes, without any post-processing. This model uses a ResNet-50 backbone, and is trained with only Cityscapes' fine data. |
| publication | Unifying Training and Inference for Panoptic Segmentation Qizhu Li, Xiaojuan Qi, Philip H.S. Torr The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2020 https://arxiv.org/abs/2001.04982 |
| project page / code | https://qizhuli.github.io/publication/unifying-training-and-inference-for-pan-seg/ |
| used Cityscapes data | fine annotations |
| used external data | ImageNet |
| runtime | n/a |
| subsampling | no |
| submission date | January, 2020 |
| previous submissions |
Average results
| Metric | All | Things | Stuff |
|---|---|---|---|
| PQ | 61.0406 | 52.7377 | 67.0791 |
| SQ | 81.4397 | 79.6345 | 82.7525 |
| RQ | 73.9323 | 66.1963 | 79.5584 |
Class results
| Class | PQ | SQ | RQ |
|---|---|---|---|
| road | 98.1999 | 98.3295 | 99.8682 |
| sidewalk | 76.0354 | 84.3132 | 90.1821 |
| building | 87.5139 | 90.3544 | 96.8562 |
| wall | 33.5078 | 75.6413 | 44.2982 |
| fence | 37.4712 | 74.4638 | 50.3214 |
| pole | 56.2795 | 68.7988 | 81.803 |
| traffic light | 56.0045 | 75.8995 | 73.7877 |
| traffic sign | 69.051 | 79.2484 | 87.1324 |
| vegetation | 89.9199 | 91.4123 | 98.3673 |
| terrain | 44.3401 | 78.7594 | 56.2982 |
| sky | 89.5468 | 93.0571 | 96.2278 |
| person | 53.951 | 77.4325 | 69.6749 |
| rider | 52.6766 | 74.4411 | 70.7629 |
| car | 64.3079 | 83.8282 | 76.714 |
| truck | 49.4698 | 85.3176 | 57.9832 |
| bus | 57.4358 | 85.0857 | 67.5035 |
| train | 52.3904 | 82.487 | 63.5135 |
| motorcycle | 47.7941 | 75.78 | 63.0695 |
| bicycle | 43.8762 | 72.704 | 60.349 |