Semantic Segmentation On Pascal Voc 2012 Val
评估指标
mIoU
评测结果
各个模型在此基准测试上的表现结果
比较表格
| 模型名称 | mIoU |
|---|---|
| rethinking-atrous-convolution-for-semantic | 82.7% |
| reliability-does-matter-an-end-to-end-weakly | 66.3 |
| fast-neural-architecture-search-of-compact | 77.1% |
| simple-does-it-weakly-supervised-instance-and | - |
| dilated-spinenet-for-semantic-segmentation | 85.64% |
| fast-neural-architecture-search-of-compact | 77.3% |
| weakly-supervised-instance-segmentation-using-1 | 53.4% |
| auto-deeplab-hierarchical-neural-architecture | 82.04% |
| hyperseg-patch-wise-hypernetwork-for-real | 80.61% |
| self-supervised-difference-detection-for-1 | 64.9 |
| find-it-if-you-can-end-to-end-adversarial | 62.8% |
| fast-neural-architecture-search-of-compact | 78.0% |
| pushing-the-limits-of-self-supervised-resnets | 77.9% |
| simple-copy-paste-is-a-strong-data | 86.6% |
| pushing-the-limits-of-self-supervised-resnets | 75.7% |
| the-devil-is-in-the-labels-semantic | 65% |
| exploiting-saliency-for-object-segmentation | 55.7% |
| pushing-the-limits-of-self-supervised-resnets | 77.3% |
| revisiting-unreasonable-effectiveness-of-data | 76.5% |
| text-image-alignment-for-diffusion-based | 87.11% |
| large-kernel-matters-improve-semantic | 81.0% |
| encoder-decoder-with-atrous-separable | - |
| exfuse-enhancing-feature-fusion-for-semantic | 85.8% |
| rethinking-pre-training-and-self-training | 90.0% |
| learning-a-discriminative-feature-network-for | 80.60% |
| deeplab-semantic-image-segmentation-with-deep | 77.69% |
| res2net-a-new-multi-scale-backbone | 79.3% |
| dicenet-dimension-wise-convolutions-for | 66.5% |
| waterfall-atrous-spatial-pooling-architecture | 80.41% |