Built independently by an author, for readers. Read the story and support ChapterPal
ChapterPal Logo
Log inSign up

keyword

compatibility learning

Compatibility learning is a machine learning approach that trains a scoring function to measure the degree of alignment or association between data representations from different feature spaces, such as input features and semantic class descriptors. Widely used in zero-shot learning frameworks, this method enables models to classify inputs from unseen categories by evaluating how well an input representation, such as an image feature vector, matches auxiliary semantic information, such as textual attributes or semantic embeddings. The scoring function can be linear, bilinear, or non-linear and is typically optimized using ranking or margin-based loss functions designed to assign higher scores to true sample-class pairs than to incorrect pairs. At inference time, an input is assigned to the class whose semantic description achieves the highest compatibility score with the input feature representation.

1 item

Zero-Shot Learning—A Comprehensive Evaluation of the Good, the Bad and the Ugly

Zero-Shot Learning—A Comprehensive Evaluation of the Good, the Bad and the Ugly

Yongqin Xian, Christoph H. Lampert, Bernt Schiele, Zeynep Akata

OrganizationsInstitute of Science and Technology AustriaMax Planck Institute for InformaticsUniversity of Amsterdam

Why you should read this

Establishes a unified evaluation benchmark and standardized data splits to resolve widespread test-set overlap in zero-shot learning, while introducing the Animals with Attributes 2 (AWA2) dataset and systematically comparing leading methods under both standard and generalized settings.

Due to the importance of zero-shot learning, i.e. classifying images where there is a lack of labeled training data, the number of proposed approaches has recently increased steadily. We argue that it is time to take a step back and to analyze the status quo of the area. The purpose of this paper is three-fold. First, given the fact that there is no agreed upon zero-shot learning benchmark, we first define a new benchmark by unifying both the evaluation protocols and data splits of publicly available datasets used for this task. This is an important contribution as published results are often not comparable and sometimes even flawed due to, e.g. pre-training on zero-shot test classes. Moreover, we propose a new zero-shot learning dataset, the Animals with Attributes 2 (AWA2) dataset which we make publicly available both in terms of image features and the images themselves. Second, we compare and analyze a significant number of the state-of-the-art methods in depth, both in the classic zero-shot setting but also in the more realistic generalized zero-shot setting. Finally, we discuss in detail the limitations of the current status of the area which can be taken as a basis for advancing it.

Added

2026-09-18