Unsupervised Domain Adaptation On Cityscapes 1

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	[email protected]
adapting-object-detectors-via-selective-cross	33.8
spatial-attention-pyramid-network-for	40.9
align-and-distill-unifying-and-improving	61.4
diversify-and-match-a-domain-adaptive	34.6
decompose-to-adapt-cross-domain-object	42.3
seeking-similarities-over-differences	43.3
align-and-distill-unifying-and-improving	59.2
awada-attention-weighted-adversarial-domain	44.8
adaptive-object-detection-with-dual-multi	38.8
improving-transferability-for-domain-adaptive	46.8
synthetic-to-real-unsupervised-domain	41.87
strong-weak-distribution-alignment-for	34.8
align-and-distill-unifying-and-improving	54.2
align-and-distill-unifying-and-improving	44.8
to-miss-attend-is-to-misalign-residual-self	43.8
align-and-distill-unifying-and-improving	62.5
self-adversarial-disentangling-for-specific	45.2
improving-object-detection-via-local-global	45.3
align-and-distill-unifying-and-improving	63.3
bi-dimensional-feature-alignment-for-cross	42.5
improving-object-detection-via-local-global	46.7
mic-masked-image-consistency-for-context	47.6
domain-adaptive-faster-r-cnn-for-object	26.1
align-and-distill-unifying-and-improving	61.7
align-and-distill-unifying-and-improving	66.8
masked-retraining-teacher-student-framework	51.2