Transductive One-Shot Learning Meet Subspace Decomposition

2025-04-01

AI Summary

Single-shot learning aims to generalize to unseen categories using only one annotated sample per novel class. This paper proposes a novel method integrating subspace decomposition with transductive inference: images are modeled as linear combinations of latent semantic primitives, and cross-category label propagation is achieved via geometric structure disentanglement. To our knowledge, this is the first work to unify transductive learning with subspace decomposition, enabling single-sample-driven discovery and transfer of semantic primitives. The approach is compatible with multiple backbone architectures—including ResNet and Vision Transformers—and achieves significant performance gains over state-of-the-art single-shot learning methods across diverse benchmark datasets. Empirical results demonstrate superior generalization capability and robustness, validating both the efficacy of the semantic decomposition paradigm and the benefits of transductive inference in extreme low-data regimes.

Technology Category

• Machine Learning: Semi-Supervised Learning
• Computer Vision: Representation Learning for Vision
• Search and Optimization: Learning to Search

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

One-shot learning focuses on adapting pretrained models to recognize newly introduced and unseen classes based on a single labeled image. While variations of few-shot and zero-shot learning exist, one-shot learning remains a challenging yet crucial problem due to its ability to generalize knowledge to unseen classes from just one human-annotated image. In this paper, we introduce a transductive one-shot learning approach that employs subspace decomposition to utilize the information from labeled images in the support set and unlabeled images in the query set. These images are decomposed into a linear combination of latent variables representing primitives captured by smaller subspaces. By representing images in the query set as linear combinations of these latent primitives, we can propagate the label from a single image in the support set to query images that share similar combinations of primitives. Through a comprehensive quantitative analysis across various neural network feature extractors and datasets, we demonstrate that our approach can effectively generalize to novel classes from just one labeled image.

Problem

• Adapt pretrained models to recognize unseen classes from one labeled image
• Utilize subspace decomposition for transductive one-shot learning
• Propagate labels by representing images as linear combinations of latent primitives

Innovation

• Transductive one-shot learning with subspace decomposition
• Linear combination of latent primitive variables
• Label propagation via shared primitive combinations