Unsupervised Projection Models
Unsupervised Projection Models are models that account for the distribution of the data in an unsupervised manner without the need manually acquired labels. They typically achieve this by using techniques that factorise the data covariance matrix or cluster related data-points into groups. These models generally exhibit a good retrieval effectiveness lying somewhere between the data independent and supervised models, but suffer from the considerable advantage of being computationally expensive at training time due to the matrix factorisation component.
| Name | Codebook | Architecture | Optimisation |
| , . |
Binary |
Shallow |
Random Hyperplanes |
| , . |
Binary |
Deep |
Backpropagation |
| Lin Chen, Hossein Esfandiari, Thomas Fu, Vahab S. Mirrokni, 2019. Locality-sensitive Hashing For F-divergences: Mutual Information Loss And Beyond |
Binary |
Shallow |
Hellinger approximation |
| , . |
Binary |
Shallow |
Variance Balance |
| , . |
Binary |
Shallow |
Matrix Factorisation |
| , . |
Binary |
Shallow |
Conditional Entropy Minimisation |
| , . |
Binary |
Shallow |
Custom Iterative Scheme |
| , . |
Binary |
Shallow |
Alternating Optimisation |
| , . |
Binary |
Shallow |
Randomised |
| , . |
Binary |
Shallow |
Matrix Factorisation |
| , . |
Binary |
Shallow |
Gradient Descent |
| , . |
Binary |
Shallow |
Variance Balance |
| , . |
Binary |
Shallow |
Matrix Factorisation |
| , . |
Binary |
Shallow |
Coordinate Descent |
| , . |
Binary |
Shallow |
Alternate Optimisation |
| , . |
Binary |
Shallow |
Random Permutations |
| , . |
Binary |
Deep |
Backpropagation |
| , . |
Binary |
Shallow |
Matrix Factorisation |
| , . |
Binary |
Shallow |
Alternative Maximisation |
| , . |
Binary |
Shallow |
Random Hyperplanes |
| , . |
Binary |
Shallow |
Augmented Lagrangian |
| , . |
Binary |
Shallow |
Random Hyperplanes |
| , . |
Binary |
Shallow |
Structured Perceptron |
| , . |
Binary |
Deep |
Backpropagation |
| Kexin Rong, Clara E. Yoon, Karianne J. Bergen, Hashem Elezabi, Peter Bailis, Philip Levis, Gregory C. Beroza, 2018. Locality-sensitive Hashing For Earthquake Detection: A Case Study Of Scaling Data-driven Science |
Binary |
Shallow |
Random Hyperplanes |
| , . |
Binary |
Deep |
Backpropagation |
| , . |
Binary |
Deep |
Backpropagation |
| , . |
Binary |
Shallow |
Random Hyperplanes |
| Jingkuan Song, Hanwang Zhang, Xiangpeng Li, Lianli Gao, Meng Wang, Richang Hong, 2018. Self-supervised Video Hashing With Hierarchical Binary Auto-encoder |
Binary |
Deep |
Backpropagation |
| , . |
Binary |
Deep |
Backpropagation |
| , . |
Binary |
Shallow |
Random Hyperplanes |
| , . |
Binary |
Shallow |
Matrix Factorisation (PCA) |
| , . |
Binary |
Shallow |
Matrix Factorisation |
| , . |
Binary |
Shallow |
Iterative optimisation |
| , . |
Binary |
Shallow |
Variance Balance |
| , . |
Binary |
Deep |
Backpropagation |
| , . |
Binary |
Deep |
Backpropagation |
| , . |
Binary |
Deep |
Backpropagation |
| , . |
Binary |
Shallow |
Fast Fourier Transform |