SOTA Video Semantic Segmentation benchmarks and papers with code

Benchmarks

Dataset Name	SOTA Method	Metric
VSPW (val)	PMT	mIoU65.7	121	3mo ago
Cityscapes (val)	Mask2Former	mIoU83.3	103	3mo ago
Cityscapes-C (test)	Time2General	mIoU49.27	78	4mo ago
VSPW	Video-kMax	mIoU63.6	55	3mo ago
CamVid	PSPNet101	mIoU76.2	41	4mo ago
NYU v2	RS-SSM	mIoU49.1	27	4mo ago
VSPW (test)	Mask2Former	mIoU56.1	25	4mo ago
VSPW 17 (test)	CFFM++	mIoU42	20	4mo ago
VIPER -> Cityscapes-Seq (val)		Road IoU96.5	14	5mo ago
Waymo	DINOv3 + GMMix	mIoU94.9	13	2mo ago
Citys s-C	Time2General	mIoU (%)43.69	12	4mo ago
Apollo	Time2General	mIoU50.51	12	4mo ago
Citys.-s (test)	Time2General	mIoU55.11	12	4mo ago
VSPW 50% warm-up ratio W2F protocol	DiTTA	mIoU52.3	9	3mo ago
VSPW W2F protocol 25% warm-up ratio	DiTTA	mIoU51	9	3mo ago
VSPW W2F protocol (10% warm-up ratio)	DiTTA	mIoU51.1	9	3mo ago
Cityscapes	TV3S	mIoU75.6	8	4mo ago
NYUDepth	TD2-PSP50	mIoU43.5	8	4mo ago
YouTube-VIS 2021	CLIP-VIS	mAP44.2	7	4mo ago
VIP-Seg (val)	Tube-Link	mIoU64.9	7	4mo ago
OVIS	CLIP-VIS	mAP18.6	6	4mo ago
Citys-s	Time2General	mIoU55.18	6	4mo ago
KITTI-360	Time2General	mIoU58.4	6	4mo ago
1024 x 512 resolution (video)	Time2General	Speed (FPS)18.15	6	4mo ago
KITTI-360 (test)	Time2General	mIoU47.71	6	4mo ago

Showing 25 of 34 rows