Review Article
Metrics Space and Norm: Taxonomy to Distance Metrics
Table 2
Literature summary of loss metrics and sample selection.
| Sample selection | Metric loss | Task |
| Easy | Triplet [107] | Recognition of image | Recognition of object | Multiple negative | Angular [101] | Clustering of image | Retrieval of image | Multiple negative | N-pair [108] | Verification of object | Clustering of image | Identification of face | Verification of face | Recognition of object | Easy | Clustering [109] | Clustering of image | Retrieval of image | Easy | Histogram [103] | Person re-Identification | Recognition of image | Retrieval of image | Hard negative | Contrastive [89] | Object recognition | Image recognition | Semi-hard negative | Quadruple [102] | Patient similarity | Easy | Part [106] | Person re-Identification | Anchor-neighbor | Hierarchical triplet [100] | Recognition of face | Retrieval of image |
|
|