The Self-Optimal-Transport Feature Transform
This addresses the need for better feature representations in computer vision and machine learning tasks, offering a flexible and efficient method that is incremental in its approach.
The paper tackles the problem of improving feature sets for downstream tasks like matching or grouping by introducing the Self-Optimal-Transport (SOT) feature transform, which encodes high-order relations and consistently enhances network performance across tasks such as unsupervised clustering, few-shot classification with state-of-the-art results, and large-scale person re-identification.
The Self-Optimal-Transport (SOT) feature transform is designed to upgrade the set of features of a data instance to facilitate downstream matching or grouping related tasks. The transformed set encodes a rich representation of high order relations between the instance features. Distances between transformed features capture their direct original similarity and their third party agreement regarding similarity to other features in the set. A particular min-cost-max-flow fractional matching problem, whose entropy regularized version can be approximated by an optimal transport (OT) optimization, results in our transductive transform which is efficient, differentiable, equivariant, parameterless and probabilistically interpretable. Empirically, the transform is highly effective and flexible in its use, consistently improving networks it is inserted into, in a variety of tasks and training schemes. We demonstrate its merits through the problem of unsupervised clustering and its efficiency and wide applicability for few-shot-classification, with state-of-the-art results, and large-scale person re-identification.