Pedestrian Detection On Citypersons

평가 지표

Heavy MR^-2

Reasonable MR^-2

Small MR^-2

Test Time

평가 결과

이 벤치마크에서 각 모델의 성능 결과

모델 이름	Heavy MR^-2	Reasonable MR^-2	Small MR^-2	Test Time	Paper Title	Repository
F2DNet (extra data)	26.23	7.8	9.43	0.44s/img	F2DNet: Fast Focal Detection Network for Pedestrian Detection
Pedestron	33.9	7.5	8.0	-	Generalizable Pedestrian Detection: The Elephant In The Room
FRCNN+Seg	-	14.8	22.6	-	CityPersons: A Diverse Dataset for Pedestrian Detection
TLL+MRF	52.0	14.4	-	-	Small-scale Pedestrian Detection Based on Somatic Topology Localization and Temporal Feature Aggregation	-
FRCNN	-	15.4	25.6	-	CityPersons: A Diverse Dataset for Pedestrian Detection
TLL	53.6	15.5	-	-	Small-scale Pedestrian Detection Based on Somatic Topology Localization and Temporal Feature Aggregation	-
LSFM	31.9	8.5	8.8	0.18	Localized Semantic Feature Mixers for Efficient Pedestrian Detection in Autonomous Driving	-
ACSP + EuroCity Persons	42.5	-	-	-	Adapted Center and Scale Prediction: More Stable and More Accurate
Beta R-CNN	47.1	10.6	-	-	Beta R-CNN: Looking into Pedestrian Detection from Another Perspective	-
NMS-Loss	-	10.08	-	-	NMS-Loss: Learning with Non-Maximum Suppression for Crowded Pedestrian Detection
DIW Loss	28.37	6.23	7.36	-	Increasing pedestrian detection performance through weighting of detection impairing factors	-
RepLoss	56.9	13.2	-	-	Repulsion Loss: Detecting Pedestrians in a Crowd
FRCNN+FPN-Res50+refined feature map+Crowdhuman	-	10.67	-	-	CrowdHuman: A Benchmark for Detecting Human in a Crowd
F2DNet	32.6	8.7	11.3	0.44s/img	F2DNet: Fast Focal Detection Network for Pedestrian Detection
ACSP	46.3	9.3	-	-	Adapted Center and Scale Prediction: More Stable and More Accurate
LSFM (Additional Data)	24.73	6.38	7.90	0.18	Localized Semantic Feature Mixers for Efficient Pedestrian Detection in Autonomous Driving	-
SOLIDER	39.4	9.7	-	-	Beyond Appearance: a Semantic Controllable Self-Supervised Learning Framework for Human-Centric Visual Tasks
CSP (with offset) + ResNet-50	49.3	11.0	16.0	0.33s/img	Center and Scale Prediction: Anchor-free Approach for Pedestrian and Face Detection
VLPD	43.1	9.4	10.9	-	VLPD: Context-Aware Pedestrian Detection via Vision-Language Semantic Self-Supervision
ALFNet	51.9	12.0	19.0	0.27	Learning Efficient Single-stage Pedestrian Detectors by Asymptotic Localization Fitting	-

0 of 22 row(s) selected.