Command Palette
Search for a command to run...
Image Clustering On Fashion Mnist
評価指標
Accuracy
NMI
評価結果
このベンチマークにおける各モデルのパフォーマンス結果
| Paper Title | |||
|---|---|---|---|
| PRCut (DinoV2) | 0.791 | 0.758 | Deep Clustering via Probabilistic Ratio-Cut Optimization |
| VMM | 0.716 | 0.710 | The VampPrior Mixture Model |
| SPC | 0.679 | 0.735 | Selective Pseudo-label Clustering |
| N2D (UMAP) | 0.672 | 0.684 | N2D: (Not Too) Deep Clustering via Clustering the Local Manifold of an Autoencoded Embedding |
| CoHiClust | 0.65 | - | Contrastive Hierarchical Clustering |
| DEN | 0.635 | 0.71 | Interpretable Visualizations with Differentiating Embedding Networks |
| PSSC | 0.628 | 0.644 | Scattering Transform Based Image Clustering using Projection onto Orthogonal Complement |
| GDL | 0.627 | 0.66 | Graph Degree Linkage: Agglomerative Clustering on a Directed Graph |
| DDC | 0.619 | 0.682 | Deep Density-based Image Clustering |
| DTI-Clustering | 0.612 | 0.637 | Deep Transformation-Invariant Clustering |
| DDC-DA | 0.609 | 0.661 | Deep Density-based Image Clustering |
| DynAE | 0.591 | 0.642 | Deep Clustering with a Dynamic Autoencoder: From Reconstruction towards Centroids Construction |
| SNNL-4 | 0.555 | 0.574 | Improving k-Means Clustering Performance with Disentangled Internal Representations |
0 of 13 row(s) selected.