List your company

Image Classification on ImageNet-1K 1.0 (test) with Performance Metrics

83.7Top-1 AccRepViT-M2.3
27.966442.435756.90571.3743May 28, 2022Sep 19, 2022Jan 12, 2023May 7, 2023Aug 30, 2023Dec 23, 2023Apr 16, 2024

Evaluation results

MethodLinks
RepViT-M2.3Type=CONV, Params (M)=22.9, GMACS=4.5, Epochs=4502023.0783.72.31,184————————
FastViT-SA36Params (M)=30.4, MACs (G)=5.62024.0483.6———241.64.3——186.5206.3138.1
EfficientFormerV2-LType=Hybrid, Params (M)=26.1, GMACS=2.6, Epochs=4502023.0783.52.7399————————
MIT-EfficientViT-B3-r224Params (M)=49.0, MACs (G)=4.02024.0483.5—————12.2——125.918.4
MNv4-Hybrid-LParams (M)=35.9, MACs (G)=7.22024.0483.4———87.63.8——61.361.818.1
EfficientFormer-L7Type=Hybrid, Params (M)=82.1, GMACS=10.2, Epochs=3002023.0783.36.6619————————
EfficientFormerV2-LType=Hybrid, Params (M)=26.1, GMACS=2.6, Epochs=3002023.0783.32.7399————————
RepViT-M2.3Type=CONV, Params (M)=22.9, GMACS=4.5, Epochs=3002023.0783.32.31,184————————
NextViT-BParams (M)=44.8, MACs (G)=8.32024.0483.2——————————
ConvNext-SParams (M)=50, MACs (G)=8.72024.0483.1———314.93.7——45.2243.918.5
SwiftFormer-L3Type=Hybrid, Params (M)=28.5, GMACS=4, Epochs=3002023.07832.91,474————————
MNv4-Conv-LParams (M)=31, MACs (G)=5.92024.0482.9———59.92.4320.837.64313.2
MIT-EfficientViT-B2-r256Params (M)=24.0, MACs (G)=2.12024.0482.7—————5.4——64.99.5
MobileViG-BType=CNN-GNN, Params (M)=26.7, GMACS=2.8, Epochs=3002023.0782.62.71,446————————
FastViT-SA24Params (M)=20.6, MACs (G)=3.82024.0482.6———171.63.22.4—131.9136.3107.5
RepViT-M1.5Type=CONV, Params (M)=14, GMACS=2.3, Epochs=4502023.0782.51.52,151————————
EfficientFormer-L3Type=Hybrid, Params (M)=31.3, GMACS=3.9, Epochs=3002023.0782.42.71,422————————
RepViT-M1.5Type=CONV, Params (M)=14, GMACS=2.3, Epochs=3002023.0782.31.52,151————————
EfficientFormerV2-S2Type=Hybrid, Params (M)=12.6, GMACS=1.3, Epochs=4502023.07821.6611————————
FastViT-SA12Type=Hybrid, Params (M)=10.9, GMACS=1.9, Epochs=3002023.0781.91.82,181————————
EfficientFormerV2-S2Type=Hybrid, Params (M)=12.6, GMACS=1.3, Epochs=3002023.0781.61.6611————————
RepViT-M1.1Type=CONV, Params (M)=8.2, GMACS=1.3, Epochs=4502023.0781.21.13,604————————
FastViT-S12Type=Hybrid, Params (M)=8.8, GMACS=1.8, Epochs=3002023.0780.91.52,313————————
SwiftFormer-L1Type=Hybrid, Params (M)=12.1, GMACS=1.6, Epochs=3002023.0780.91.62,576————————
RepViT-M1.1Type=CONV, Params (M)=8.2, GMACS=1.3, Epochs=3002023.0780.71.13,604————————
MNv4-Hybrid-MParams (M)=10.5, MACs (G)=1.22024.0480.7———14.31.5——17.910.85.9
MobileViG-MType=CNN-GNN, Params (M)=14, GMACS=1.5, Epochs=3002023.0780.61.62,491————————
FastViT-SA12Params (M)=10.9, MACs (G)=1.92024.0480.6———86.521.6—79.669.552.1
RepViT-M1.0Type=CONV, Params (M)=6.8, GMACS=1.1, Epochs=4502023.0780.313,910————————
RepViT-M1.0Type=CONV, Params (M)=6.8, GMACS=1.1, Epochs=3002023.078013,910————————
MNv4-Conv-MParams (M)=9.2, MACs (G)=1.02024.0479.9———11.41.11.17.318.18.64.1
FastViT-S12Params (M)=8.8, MACs (G)=1.82024.0479.8———831.81.6—7569.247
EfficientFormerV2-S1Type=Hybrid, Params (M)=6.1, GMACS=0.7, Epochs=4502023.0779.71.11,153————————
MobileOne-S4Params (M)=14.8, MACs (G)=1.52024.0479.4———26.71.71.5928.619.45.9
MIT-EfficientViT-B1-r224Params (M)=9.1, MACs (G)=0.52024.0479.4—————2.4——18.15
EfficientFormer-L1Type=Hybrid, Params (M)=12.3, GMACS=1.3, Epochs=3002023.0779.21.43,360————————
RepViT-M0.9Type=CONV, Params (M)=5.1, GMACS=0.8, Epochs=4502023.0779.10.94,817————————
EfficientFormerV2-S1Type=Hybrid, Params (M)=6.1, GMACS=0.7, Epochs=3002023.07791.11,153————————
RepViT-M0.9Type=CONV, Params (M)=5.1, GMACS=0.8, Epochs=3002023.0778.70.94,817————————
SwiftFormer-SType=Hybrid, Params (M)=6.1, GMACS=1, Epochs=3002023.0778.51.23,376————————
MobileNet-V2-2.0xParams (M)=11.2, MACs (G)=1.12024.0478.4———13.91.11.56.919.110.64.2
MobileViG-SType=CNN-GNN, Params (M)=7.2, GMACS=1, Epochs=3002023.0778.21.22,985————————
MultiHardware-AVG-1.5xParams (M)=10.0, MACs (G)=1.02024.0478.2———121.1—6.120.3—4.5
MultiHardware-MAX-1.5xParams (M)=8.9, MACs (G)=0.82024.0477.9———9.81—5.723.2—4.1
MobileNet-V2-1.5xParams (M)=6.8, MACs (G)=0.72024.0476.8———9.30.915.616.47.33.3
FastViT-T8Type=Hybrid, Params (M)=3.6, GMACS=0.7, Epochs=3002023.0776.70.93,909————————
EfficientFormerV2-S0Type=Hybrid, Params (M)=3.5, GMACS=0.4, Epochs=4502023.0776.20.91,274————————
MobileViG-TiType=CNN-GNN, Params (M)=5.2, GMACS=0.7, Epochs=3002023.0775.714,337————————
SwiftFormer-XSType=Hybrid, Params (M)=3.5, GMACS=0.6, Epochs=3002023.0775.714,304————————
EfficientFormerV2-S0Type=Hybrid, Params (M)=3.5, GMACS=0.4, Epochs=3002023.0775.70.91,274————————
FastViT-T8+Params (M)=3.6, MACs (G)=0.72024.0475.6———49.31.30.7—40.743.624.7
MobileNet-V1Params (M)=4.2, MACs (G)=0.62024.0474———6.10.80.73.2134.62.1
MNv4-Conv-SParams (M)=3.8, MACs (G)=0.22024.0473.8———2.40.70.62.48.41.82
MobileNet-V2Params (M)=3.5, MACs (G)=0.32024.0473.4———50.70.73.913.64.12.5
MobileOne-S0Params (M)=2.1, MACs (G)=0.32024.0471.4———4.20.70.52.910.73.31.7
WaveMix-240/12 (3 level)# Param.=33.8 M, GPU RAM for batch size 64=15.6 GB, Input Resolution=224x224, Data Augmentation=None2022.0570.02—610205———————
MobileNet-V3L-0.5xParams (M)=2.7, MACs (G)=0.12024.0469.2———2.40.80.453.59.922.1
MobileNet-V2-0.5xParams (M)=2.0, MACs (G)=0.12024.0466———2.40.70.52.98.31.81.9
WaveMix-224/12 (2 level)# Param.=22.6 M, GPU RAM for batch size 64=9.8 GB, Input Resolution=224x224, Data Augmentation=None2022.0565.9—1,250385———————
ResNet-152# Param.=60.2 M, GPU RAM for batch size 64=12.7 GB, Input Resolution=224x224, Data Augmentation=None2022.0565.86—758344———————
ResNet-101# Param.=44.5 M, GPU RAM for batch size 64=9.6 GB, Input Resolution=224x224, Data Augmentation=None2022.0564.6—725487———————
ConvMixer-1024/12# Param.=14.6 M, GPU RAM for batch size 64=23.6 GB, Input Resolution=224x224, Data Augmentation=None2022.0564.13—667251———————
ConvMixer-512/16# Param.=5.4 M, GPU RAM for batch size 64=14.1 GB, Input Resolution=224x224, Data Augmentation=None2022.0562.24—725220———————
ResNet-50# Param.=25.6 M, GPU RAM for batch size 64=6.2 GB, Input Resolution=224x224, Data Augmentation=None2022.0561.76—617638———————
ConvMixer-512/12# Param.=4.2 M, GPU RAM for batch size 64=10.8 GB, Input Resolution=224x224, Data Augmentation=None2022.0560.24—735292———————
MobileNetv3-large# Param.=5.5 M, GPU RAM for batch size 64=3.5 GB, Input Resolution=224x224, Data Augmentation=None2022.0558.89—481492———————
ResNet-34# Param.=21.8 M, GPU RAM for batch size 64=3.1 GB, Input Resolution=224x224, Data Augmentation=None2022.0557.02—410414———————
ResNet-18# Param.=11.7 M, GPU RAM for batch size 64=2.7 GB, Input Resolution=224x224, Data Augmentation=None2022.0555.12—439450———————
WaveMix-Lite-128/8# Param.=3.9 M, GPU RAM for batch size 64=4.5 GB, Input Resolution=224x224, Data Augmentation=None2022.0554.12—1,7241,242———————
MobileNetv3-small# Param.=2.5 M, GPU RAM for batch size 64=1.4 GB, Input Resolution=224x224, Data Augmentation=None2022.0551.57—229255———————
ViT-B-16# Param.=86.6 M, GPU RAM for batch size 64=10.0 GB, Input Resolution=224x224, Data Augmentation=None2022.0539.53—420140———————
ViT-B-32# Param.=88.2 M, GPU RAM for batch size 64=2.2 GB, Input Resolution=224x224, Data Augmentation=None2022.0530.11—1,6131,595———————