Crowd Counting On Shanghaitech B

评估指标

MAE

评测结果

各个模型在此基准测试上的表现结果

模型名称	MAE	Paper Title	Repository
ic-CNN	10.7	Iterative Crowd Counting	-
Cascaded-MTL	20	CNN-based Cascaded Multi-task Learning of High-level Prior and Density Estimation for Crowd Counting
CLIP-EBC (ViT-B/16)	6.6	CLIP-EBC: CLIP Can Count Accurately through Enhanced Blockwise Classification
GauNet (ResNet-50)	6.0	Rethinking Spatial Invariance of Convolutional Networks for Object Counting
SAFECount	9.98	Few-shot Object Counting with Similarity-Aware Feature Enhancement
Liu et al.	13.7	Leveraging Unlabeled Data for Crowd Counting by Learning to Rank
CLIP-EBC (ResNet50)	6.0	CLIP-EBC: CLIP Can Count Accurately through Enhanced Blockwise Classification
APGCC	8.7	Improving Point-based Crowd Counting and Localization Based on Auxiliary Point Guidance
Switch-CNN	21.6	Switching Convolutional Neural Network for Crowd Counting
CSRNet	10.6	CSRNet: Dilated Convolutional Neural Networks for Understanding the Highly Congested Scenes
OrdinalEntropy	9.1	Improving Deep Regression with Ordinal Entropy
S-DCNet	6.7	From Open Set to Closed Set: Counting Objects by Spatial Divide-and-Conquer	-
M-SFANet+M-SegNet	6.32	Encoder-Decoder Based Convolutional Neural Networks with Multi-Scale-Aware Modules for Crowd Counting
Zhang et al.	32.0	Cross-Scene Crowd Counting via Deep Convolutional Neural Networks	-
CSRNet-EBC	6.9	CLIP-EBC: CLIP Can Count Accurately through Enhanced Blockwise Classification
DMCount-EBC	7.0	CLIP-EBC: CLIP Can Count Accurately through Enhanced Blockwise Classification
FusionCount	6.9	FusionCount: Efficient Crowd Counting via Multiscale Feature Fusion
DM-Count	7.4	Distribution Matching for Crowd Counting
IG-CNN	13.6	Divide and Grow: Capturing Huge Diversity in Crowd Images with Incrementally Growing CNN	-
MCNN	26.4	Single-Image Crowd Counting via Multi-Column Convolutional Neural Network

0 of 31 row(s) selected.