Collaborating Authors

Google's TensorFlow Similarity helps AI models find related items


The Transform Technology Summits start October 13th with Low-Code/No Code: Enabling Enterprise Agility. Google today announced TensorFlow Similarity, a Python package designed to train similarity models with the company's TensorFlow machine learning framework. Similarity models search for related items, for example finding similar-looking clothes and identifying currently playing songs. As Google explains, many similarity models are trained using a technique called contrastive learning. Contrastive learning, in turn, relies on clustering algorithms, which automatically identify patterns in data by operating on the theory that data points in groups should have similar features.

Discriminative Similarity for Clustering and Semi-Supervised Learning Machine Learning

Similarity-based clustering and semi-supervised learning methods separate the data into clusters or classes according to the pairwise similarity between the data, and the pairwise similarity is crucial for their performance. In this paper, we propose a novel discriminative similarity learning framework which learns discriminative similarity for either data clustering or semi-supervised learning. The proposed framework learns classifier from each hypothetical labeling, and searches for the optimal labeling by minimizing the generalization error of the learned classifiers associated with the hypothetical labeling. Kernel classifier is employed in our framework. By generalization analysis via Rademacher complexity, the generalization error bound for the kernel classifier learned from hypothetical labeling is expressed as the sum of pairwise similarity between the data from different classes, parameterized by the weights of the kernel classifier. Such pairwise similarity serves as the discriminative similarity for the purpose of clustering and semi-supervised learning, and discriminative similarity with similar form can also be induced by the integrated squared error bound for kernel density classification. Based on the discriminative similarity induced by the kernel classifier, we propose new clustering and semi-supervised learning methods.

Deconfounded Representation Similarity for Comparison of Neural Networks Machine Learning

Similarity metrics such as representational similarity analysis (RSA) and centered kernel alignment (CKA) have been used to compare layer-wise representations between neural networks. However, these metrics are confounded by the population structure of data items in the input space, leading to spuriously high similarity for even completely random neural networks and inconsistent domain relations in transfer learning. We introduce a simple and generally applicable fix to adjust for the confounder with covariate adjustment regression, which retains the intuitive invariance properties of the original similarity measures. We show that deconfounding the similarity metrics increases the resolution of detecting semantically similar neural networks. Moreover, in real-world applications, deconfounding improves the consistency of representation similarities with domain similarities in transfer learning, and increases correlation with out-of-distribution accuracy.

Object Similarity by Humans and Machines

AAAI Conferences

In this paper, we briefly address a research regarding how to objectively evaluate machine-based object similarity measures by human-based estimation. Based on a novel approach for similarity measure of 3-D objects we create a ground truth of 3-D objects and their similarities estimated by humans. The automatic similarity results achieved are evaluated against this ground truth in terms of precision and recall in an object retrieval scenario. To further illustrate the reciprocity properties between machine and human perception, we compare the similarities achieved by both on testing data and show how it can be used to address other problems and formulations.

An Online Algorithm for Large Scale Image Similarity Learning

Neural Information Processing Systems

Learning a measure of similarity between pairs of objects is a fundamental problem in machine learning. It stands in the core of classification methods like kernel machines, and is particularly useful for applications like searching for images that are similar to a given image or finding videos that are relevant to a given video. In these tasks, users look for objects that are not only visually similar but also semantically related to a given object. Unfortunately, current approaches for learning similarity may not scale to large datasets with high dimensionality, especially when imposing metric constraints on the learned similarity. We describe OASIS, a method for learning pairwise similarity that is fast and scales linearly with the number of objects and the number of non-zero features.