Clustering on USPS (test)
82.55ACCDnC-SC
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| DnC-SC2021.04 | 82.55 | — | — | |
| U-SPEC2021.04 | 80.79 | — | — | |
| IEC2017.03 | 76.7 | 64.1 | — | |
| DBC2017.03 | 74.3 | 72.4 | — | |
| LSC-K2021.04 | 74.02 | — | — | |
| DCNinterpretable hard assignments=true, end-to-end learning=false2019.10 | 73.9 | 71.9 | 61.9 | |
| LSC-R2021.04 | 73.9 | — | — | |
| LSC-RH2021.04 | 73.89 | — | — | |
| LSC-KH2021.04 | 73.66 | — | — | |
| SC2021.04 | 73.21 | — | — | |
| CKMinterpretable hard assignments=true, end-to-end learning=true2019.10 | 72.1 | 70.7 | 61.3 | |
| AEC2017.03 | 71.5 | 65.1 | — | |
| DECinterpretable hard assignments=false, end-to-end learning=true2019.10 | 71.1 | 72.6 | 63.8 | |
| Nyström2021.04 | 69.47 | — | — | |
| AE+KMinterpretable hard assignments=true, end-to-end learning=false2019.10 | 68.4 | 68.1 | 59.4 | |
| KM2021.04 | 67.01 | — | — | |
| FCAE-KMS2017.03 | 66.7 | 64.5 | — | |
| KMinterpretable hard assignments=true, end-to-end learning=false2019.10 | 58 | 60 | 44 | |
| KMS2017.03 | 53.5 | 53.1 | — | |
| AggloBackbone=No-AE2026.01 | — | — | 50.8 | |
| AggloBackbone=Conv-AE2026.01 | — | — | 58.7 | |
| ClAMBackbone=No-AE2026.01 | — | — | 47.3 | |
| ClAMBackbone=Conv-AE2026.01 | — | — | 52.7 | |
| DCAMBackbone=Conv-AE2026.01 | — | — | 66.8 | |
| DCAMBackbone=EDCWRN-AE2026.01 | — | — | 68.4 | |
| DCAMBackbone=Resnet-AE2026.01 | — | — | 68.9 | |
| DCECBackbone=Conv-AE2026.01 | — | — | 65.4 | |
| DCECBackbone=Resnet-AE2026.01 | — | — | 67.3 | |
| EDCWRN2026.01 | — | — | 63.7 | |
| KmeansBackbone=No-AE2026.01 | — | — | 47.6 | |
| KmeansBackbone=Conv-AE2026.01 | — | — | 54.1 |