HyperAI초신경

홈 뉴스 최신 연구 논문 튜토리얼 데이터셋 백과사전 SOTA LLM 모델 GPU 랭킹 컨퍼런스

한국어

HyperAI초신경

Depth Completion On Kitti Depth Completion

평가 지표

MAE

RMSE

Runtime [ms]

평가 결과

이 벤치마크에서 각 모델의 성능 결과

모델 이름	MAE	RMSE	Runtime [ms]	Paper Title	Repository
SparseConvs	481	1601	10	Sparsity Invariant CNNs
NConv-CNN-L2	233	830	20	Confidence Propagation through CNNs for Guided Sparse Depth Regression
Spade-sD	248	1035	40	Sparse and Dense Data with CNNs: Depth Completion and Semantic Segmentation	-
NLSPN	199.59	741.68	220	Non-Local Spatial Propagation Network for Depth Completion
SemAttNet	-	709.41	-	SemAttNet: Towards Attention-based Semantic Aware Guided Depth Completion
PENet	210.55	730.08	32	PENet: Towards Precise and Efficient Image Guided Depth Completion
RigNet	-	713.44	-	RigNet: Repetitive Image Guided Network for Depth Completion	-
NConv-CNN	360	1268	10	Confidence Propagation through CNNs for Guided Sparse Depth Regression
VOICED	299.41	1169.97	20	Unsupervised Depth Completion from Visual Inertial Odometry
Spade-RGBsD	235	918	70	Sparse and Dense Data with CNNs: Depth Completion and Semantic Segmentation	-
KBNet	256.76	1069.47	16	Unsupervised Depth Completion with Calibrated Backprojection Layers
FusionNet (RGB_guide&certainty)	215.02	772.87	20	Sparse and noisy LiDAR completion with RGB guidance and uncertainty
DVMN	220.37	776.31	-	DVMN: Dense Validity Mask Network for Depth Completion	-
ScaffNet-FusionNet	280.76	1121.93	-	Learning Topology from Synthetic Data for Unsupervised Depth Completion
HMS-Net	258	937	-	HMS-Net: Hierarchical Multi-scale Sparsity-invariant Network for Sparse Depth Completion	-
NConv-CNN-L1	208	859	20	Confidence Propagation through CNNs for Guided Sparse Depth Regression

0 of 16 row(s) selected.