Semantic Segmentation On Cityscapes 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 |
|---|---|
| exploring-semantic-segmentation-on-the-dct | 61.6 |
| improve-vision-transformers-training-by | 83.6% |
| efficient-visual-pretraining-with-contrastive | 77.0% |
| streaming-multiscale-deep-equilibrium-models | 78.2 |
| rethinking-semantic-segmentation-from-a | 82.15 |
| soft-labelling-for-semantic-segmentation | 84.8 |
| fasterseg-searching-for-faster-real-time-1 | 73.1% |
| 190807919 | 80.2 |
| streaming-multiscale-deep-equilibrium-models | 57.9 |
| deep-residual-learning-for-image-recognition | 75.7 |
| vpnext-rethinking-dense-decoding-for-plain | 84.4 |
| gated-scnn-gated-shape-cnns-for-semantic | 73.0% |
| understanding-the-robustness-in-vision | 82.3 |
| efficientvit-enhanced-linear-attention-for | 83.2 |
| trans4trans-efficient-transformer-for-1 | 81.54% |
| aerial-pass-panoramic-annular-scene | 72.8% |
| dsnet-a-novel-way-to-use-atrous-convolutions | 80.4 |
| repvgg-making-vgg-style-convnets-great-again | 80.57% |
| internimage-exploring-large-scale-vision | 87 |
| vision-transformer-adapter-for-dense | 85.8 |
| template-based-automatic-search-of-compact | 68.1% |
| wavemix-lite-a-resource-efficient-neural | 82.60 |
| multiscale-deep-equilibrium-models | 80.3% |
| csfnet-a-cosine-similarity-fusion-network-for | 76.36 |
| contextnet-exploring-context-and-detail-for | 65.9% |
| unet-a-nested-u-net-architecture-for-medical | 75.5 |
| oneformer-one-transformer-to-rule-universal | 84.6 |
| swinmtl-a-shared-architecture-for | 76.41 |
| oneformer-one-transformer-to-rule-universal | 84.4 |
| eeea-net-an-early-exit-evolutionary-neural | 76.8 |
| polarized-self-attention-towards-high-quality-1 | 86.93 |
| segformer-simple-and-efficient-design-for | 84.0 |
| fully-attentional-networks-with-self-emerging-1 | 82.8 |
| encoder-decoder-with-atrous-separable | 79.6 |
| mrfp-learning-generalizable-semantic | 42.4 |
| hyperbolic-active-learning-for-semantic | 77.8 |
| masked-attention-mask-transformer-for | 84.3 |
| 模型 38 | 81.5 |
| wavemix-lite-a-resource-efficient-neural | 82.7 |
| conditional-boundary-loss-for-semantic | 83.4 |
| semask-semantically-masked-transformers-for-1 | 80.39 |
| hierarchical-multi-scale-attention-for | 86.3 |
| pixel-wise-anomaly-detection-in-complex | 83.5 |
| segformer-simple-and-efficient-design-for | - |
| harnessing-diffusion-models-for-visual | 87.1 |
| squeezenas-fast-neural-architecture-search | 68.0% |
| internimage-exploring-large-scale-vision | 86.4 |
| volo-vision-outlooker-for-visual-recognition | 84.3 |
| depth-anything-unleashing-the-power-of-large | 86.2 |
| streaming-multiscale-deep-equilibrium-models | 45.5 |
| auto-deeplab-hierarchical-neural-architecture | 80.33% |
| pushing-the-limits-of-self-supervised-resnets | 74.6 |
| sequential-ensembling-for-semantic | 84.8 |
| template-based-automatic-search-of-compact | 69.5% |
| multiscale-deep-equilibrium-models | 77.8% |
| waterfall-atrous-spatial-pooling-architecture | 74% |
| gated-scnn-gated-shape-cnns-for-semantic | 74.7% |
| panoptic-deeplab-a-simple-strong-and-fast | 81.5% |
| dilated-neighborhood-attention-transformer | 84.5 |
| semask-semantically-masked-transformers-for-1 | 84.98 |
| streaming-multiscale-deep-equilibrium-models | 71.5 |
| squeezenas-fast-neural-architecture-search | 73.6% |
| cmx-cross-modal-fusion-for-rgb-x-semantic | 82.6 |
| squeezenas-fast-neural-architecture-search | 75.2% |
| ds-pass-detail-sensitive-panoramic-annular | 72.1% |
| dilated-spinenet-for-semantic-segmentation | 83.04% |
| real-time-fusion-network-for-rgb-d-semantic | 72.5% |
| sernet-former-semantic-segmentation-by | 87.35 |
| rethinking-decoders-for-transformer-based | 78.8 |
| oneformer-one-transformer-to-rule-universal | 85.8 |
| mrfp-learning-generalizable-semantic | 34.66 |
| pushing-the-limits-of-self-supervised-resnets | 75.2 |
| rethinking-decoders-for-transformer-based | 81.0 |
| object-contextual-representations-for | 83.6 |
| erfnet-efficient-residual-factorized-convnet | 72.1% |
| wavemix-lite-a-resource-efficient-neural-1 | 76.79 |
| pyramid-scene-parsing-network | 79.7 |
| beyond-self-attention-external-attention | 81.7% |
| hrvit-multi-scale-high-resolution-vision | 82.81% |
| bending-reality-distortion-aware-transformers | 81.1% |
| repmlpnet-hierarchical-vision-mlp-with-re | 76.27 |
| dsnet-a-novel-way-to-use-atrous-convolutions | 82.0 |
| standardized-max-logits-a-simple-yet | 80.33 |
| ddp-diffusion-model-for-dense-visual | 83.9 |
| fast-scnn-fast-semantic-segmentation-network | 69.19 |
| dicenet-dimension-wise-convolutions-for | 63.4 |
| cmx-cross-modal-fusion-for-rgb-x-semantic | 81.6 |
| pointrend-image-segmentation-as-rendering | 78.6 |
| bending-reality-distortion-aware-transformers | 79.1% |
| object-contextual-representations-for | 80.6 |
| rethinking-atrous-convolution-for-semantic | 78.5% |
| incorporating-luminance-depth-and-color | 68.48% |
| csfnet-a-cosine-similarity-fusion-network-for | 74.73 |
| 190807919 | 81.1 |
| resnest-split-attention-networks | 82.7 |
| hrvit-multi-scale-high-resolution-vision | 83.16% |
| hrvit-multi-scale-high-resolution-vision | 81.63% |