Object Counting on CARPK

Metric: RMSE (lower is better)

LeaderboardDataset

Loading chart...

Results

Sort:

#	Model↕	RMSE▲	Extra Data	Paper	Date↕	Code
1	HLCNN	3.02	No	-	-	Code
2	CLIP-LOCAR	6.02	No	Car Object Counting and Position Estimation via ...	2025-07-11	Code
3	SAFECount	7.04	No	Few-shot Object Counting with Similarity-Aware F...	2022-01-22	Code
4	CounTR	7.45	No	CounTR: Transformer-based Generalised Visual Cou...	2022-08-29	Code
5	BMNet+	7.83	No	Represent, Compare, and Learn: A Similarity-Awar...	2022-03-16	Code
6	Soft-IoU + EM-Merger unit	8.52	No	Precise Detection in Densely Packed Scenes	2019-04-01	Code
7	VLCounter	8.68	No	VLCounter: Text-aware Visual Representation for ...	2023-12-27	Code
8	CounTX (uses arbitrary text input to specify object to count, used "the cars" for CARPK)	10.87	No	Open-world Text-specified Object Counting	2023-06-02	Code
9	RetinaNet (2018)	22.3	No	Drone-based Object Counting by Spatially Regular...	2017-07-19	-
10	LPN Counting (2017)	34.46	No	Drone-based Object Counting by Spatially Regular...	2017-07-19	-
11	One-Look Regression (2016)	36.73	No	A Large Contextual Dataset for Classification, D...	2016-09-14	-
12	Faster R-CNN (2015)	47.67	No	Faster R-CNN: Towards Real-Time Object Detection...	2015-06-04	Code
13	YOLO (2016)	57.55	No	You Only Look Once: Unified, Real-Time Object De...	2015-06-08	Code
14	YOLO9000opt (2017)	172.46	No	YOLO9000: Better, Faster, Stronger	2016-12-25	Code

#1HLCNN
3.02
RMSE
No paperCode
#2CLIP-LOCARSOTA
6.02
RMSE· 2025-07-11
Car Object Counting and Position Estimation via Extension of the CLIP-EBC Framework Code
#3SAFECountSOTA
7.04
RMSE· 2022-01-22
Few-shot Object Counting with Similarity-Aware Feature Enhancement Code
#4CounTR
7.45
RMSE· 2022-08-29
CounTR: Transformer-based Generalised Visual Counting Code
#5BMNet+
7.83
RMSE· 2022-03-16
Represent, Compare, and Learn: A Similarity-Aware Framework for Class-Agnostic Counting Code
#6Soft-IoU + EM-Merger unitSOTA
8.52
RMSE· 2019-04-01
Precise Detection in Densely Packed Scenes Code
#7VLCounter
8.68
RMSE· 2023-12-27
VLCounter: Text-aware Visual Representation for Zero-Shot Object Counting Code
#8CounTX (uses arbitrary text input to specify object to count, used "the cars" for CARPK)
10.87
RMSE· 2023-06-02
Open-world Text-specified Object Counting Code
#9RetinaNet (2018)SOTA
22.3
RMSE· 2017-07-19
Drone-based Object Counting by Spatially Regularized Regional Proposal Network
#10LPN Counting (2017)SOTA
34.46
RMSE· 2017-07-19
Drone-based Object Counting by Spatially Regularized Regional Proposal Network
#11One-Look Regression (2016)SOTA
36.73
RMSE· 2016-09-14
A Large Contextual Dataset for Classification, Detection and Counting of Cars with Deep Learning
#12Faster R-CNN (2015)SOTA
47.67
RMSE· 2015-06-04
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks Code
#13YOLO (2016)
57.55
RMSE· 2015-06-08
You Only Look Once: Unified, Real-Time Object Detection Code
#14YOLO9000opt (2017)
172.46
RMSE· 2016-12-25
YOLO9000: Better, Faster, Stronger Code