Table 7 OA results across various methods on the AID30 dataset.

From: Pure data correction enhancing remote sensing image classification with a lightweight ensemble model

Method	Technique feature	Params (M)	Published year	OA (%)
Method	Technique feature	Params (M)	Published year	TR-20%	TR-50%
SLGE-CNN⁷	Single CNN	5.1	TGRS2022	96.10 ± 0.18 (TR-80%)
AF-CNN⁸		3.8	TGRS2022	95.96 (TR-60%)
LCNN-HWCF¹¹		0.6	RS2022	95.76 ± 0.16	97.43 ± 0.28
SCCNN⁴²		0.5	RS2022	93.15 ± 0.25	97.31 ± 0.10
MF2CNet⁴³		33.2	TGRS2022	95.54 ± 0.17	97.02 ± 0.28
ViT-Base⁴⁴	Single ViT	86.6	RS2021	95.86 ± 0.28	None
ET-GSNet¹²	Single ViT	98.3	TGRS2022	95.58 ± 0.18	96.88 ± 0.19
ViTAEv2¹³	Single Swin-T	27.6	TGRS2023	96.81 ± 0.03	98.30 ± 0.04
EfficientNet-B3-Attn⁴⁵	Attention module for CNN	> 12.0	ACCESS2021	94.45 ± 0.76	96.56 ± 0.12
MBLANet⁴⁶		> 25.6	TIP2022	95.60 ± 0.17	97.14 ± 0.03
EAM-CNN¹⁴		> 46.8	GRSL2021	94.26 ± 0.11	97.06 ± 0.19
LHNet⁴⁷	Feature fusion	> 46.8	TGRS2022	93.30 ± 0.10	97.81 ± 0.13
SCViT¹⁵		40.1	TGRS2022	95.56 ± 0.17	96.98 ± 0.16
MLF2Net⁴⁸		23.8	GRSL2022	95.44 ± 0.25	97.08 ± 0.17
SEMSDNet⁴⁹		3.7	JSTARS2021	94.23 ± 0.63	97.64 ± 0.51
LmNet⁵⁰		> 25.0	ACESS2021	95.82 ± 0.25	97.12 ± 0.14
D-CNN⁵¹	Multiple models	None	RS2021	94.63	96.43
GCSANet¹⁶		8.1	JSTARS2022	95.96 ± 0.38	97.53 ± 0.32
ACGLNet¹⁷		33.6	RS2022	94.44 ± 0.09	96.10 ± 0.10
SF-MSFormer¹⁸		36.3	TGRS2023	None	98.72 ± 0.31
AGOS⁵²		None	TGRS2022	95.81 ± 0.25	97.43 ± 0.21
GRMA-Net¹⁹		54.1	TGRS2022	96.19 ± 0.48	97.84 ± 0.39
ACNet⁵³		> 276.6	JSTARS2021	93.33 ± 0.29	95.38 ± 0.29
T-CNN⁵⁴		15.9	TGRS2022	94.55 ± 0.27	96.72 ± 0.23
GLDBS⁵⁵		> 23.4	GRSL2022	95.45 ± 0.19	97.01 ± 0.22
TRS⁵⁶		46.3	RS2021	95.54 ± 0.18	98.48 ± 0.06
CTNet⁵⁷		> 107.8	GRSL2022	96.25 ± 0.10	97.70 ± 0.11
HHTL⁵⁸		> 173.2	JSTARS2022	95.62 ± 0.13	96.88 ± 0.21
L2RCF⁵⁹	Custom learning framework	46.7	TGRS2023	97.00 ± 0.17	97.80 ± 0.22
ViT-CL⁶⁰		86.0	JSTARS2023	95.60	97.42
GSCCTL⁶¹		None	IJRS2022	91.32	None
MGDNet⁶²		None	TGRS2023	86.52 ± 0.81	None
TSTNet²¹	Multiple Swin-Ts	173.0	RS2002	97.20 ± 0.22	98.70 ± 0.12
IBSwin-CR²²		164.0	JSTARS2023	97.61 ± 0.12	98.78 ± 0.09
MFST⁶³		None	GRSL2022	96.23 ± 0.16	97.38 ± 0.08
mmsCNN-HMM²⁷	Multi-CNN ensemble	19.0	RS2022	93.93 ± 0.15	97.81 ± 0.04
MGML²⁸		None	TNNLS2023	94.47 ± 0.15	97.89 ± 0.07
ESD-MBENet²⁹		23.9	TGRS2022	96.39 ± 0.21	98.40 ± 0.23
RC-B3	Single CNN	12.2	This work	97.22 ± 0.03	98.27 ± 0.04
Swin-T-Tiny	Single ViT	28.3		97.42 ± 0.10	98.39 ± 0.11
N-ViT-S	Single ViT	31.7		97.62 ± 0.11	98.47 ± 0.02
RD-ESE	Dual-CNN ensemble	17.5		97.53 ± 0.02	98.45 ± 0.02
RST-ESE	CNN-ViT hybrid ensemble	40.5		97.86 ± 0.12	98.59 ± 0.05
RNV-ESE	CNN-ViT hybrid ensemble	43.9		97.86 ± 0.12	98.59 ± 0.15

Back to article page

Table 7 OA results across various methods on the AID30 dataset.

Search

Quick links