Semantic Segmentation On Bdd100K Val
評価指標
mIoU
評価結果
このベンチマークにおける各モデルのパフォーマンス結果
比較表
| モデル名 | mIoU |
|---|---|
| encoder-decoder-with-atrous-separable | 63.6 |
| dsnet-a-novel-way-to-use-atrous-convolutions | 64.6 |
| partial-order-pruning-for-best-speedaccuracy | 42.5(82.3fps) |
| sernet-former-semantic-segmentation-by | 67.42 |
| fast-and-accurate-scene-parsing-via-bi | 53.4(42.1fps) |
| expectation-maximization-attention-networks | 61.4 |
| bisenet-bilateral-segmentation-network-for | 53.8(45.1fps) |
| object-contextual-representations-for | 60.1 |
| pyramid-scene-parsing-network | 62.3 |
| sfnet-faster-accurate-and-domain-agnostic | 60.6(194.5FPS 4090) |
| icnet-for-real-time-semantic-segmentation-on | 52.4(39.5fps) |
| partial-order-pruning-for-best-speedaccuracy | 47.8(53.4fps) |
| what-s-there-in-the-dark | 53.52 |
| semantic-flow-for-fast-and-accurate-scene | 60.6(132.5FPS 4090) |
| vltseg-simple-transfer-of-clip-based-vision | 72.5 |
| semantic-flow-for-fast-and-accurate-scene | 55.4(70.3fps) |
| rethinking-bisenet-for-real-time-semantic | 52.1(45.8FPS) |
| sfnet-faster-accurate-and-domain-agnostic | 60.6(161.3FPS 4090) |
| dsnet-a-novel-way-to-use-atrous-convolutions | 62.6(172.2FPS 4090) |
| rethinking-bisenet-for-real-time-semantic | 53.8(33.0FPS) |
| mrfp-learning-generalizable-semantic | 39.55 |
| dual-attention-network-for-scene-segmentation | 62.8 |
| mrfp-learning-generalizable-semantic | 31.44 |
| semantic-flow-for-fast-and-accurate-scene | 60.2(208FPS 4090) |