HyperAI초신경

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

한국어

HyperAI초신경

Object Counting On Carpk

평가 지표

MAE

RMSE

평가 결과

이 벤치마크에서 각 모델의 성능 결과

모델 이름	MAE	RMSE	Paper Title	Repository
YOLO (2016)	156.00	57.55	You Only Look Once: Unified, Real-Time Object Detection
Faster R-CNN (2015)	39.88	47.67	Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
One-Look Regression (2016)	21.88	36.73	A Large Contextual Dataset for Classification, Detection and Counting of Cars with Deep Learning	-
Soft-IoU + EM-Merger unit	6.77	8.52	Precise Detection in Densely Packed Scenes
LPN Counting (2017)	22.76	34.46	Drone-based Object Counting by Spatially Regularized Regional Proposal Network	-
HLCNN	2.12	3.02	An Accurate Car Counting in Aerial Images Based on Convolutional Neural Networks
VLCounter	6.46	8.68	VLCounter: Text-aware Visual Representation for Zero-Shot Object Counting
CounTR	5.75	7.45	CounTR: Transformer-based Generalised Visual Counting
SAFECount	5.33	7.04	Few-shot Object Counting with Similarity-Aware Feature Enhancement
YOLO9000opt (2017)	130.40	172.46	YOLO9000: Better, Faster, Stronger
CounTX (uses arbitrary text input to specify object to count, used "the cars" for CARPK)	8.13	10.87	Open-world Text-specified Object Counting
RetinaNet (2018)	16.62	22.30	Drone-based Object Counting by Spatially Regularized Regional Proposal Network	-
RetinaNet (2018)	24.58	-	Focal Loss for Dense Object Detection
BMNet+	5.76	7.83	Represent, Compare, and Learn: A Similarity-Aware Framework for Class-Agnostic Counting

0 of 14 row(s) selected.