Locality and compositionality in zero-shot learning

2019/12/20 by Tristan Sylvain, Linda Petrini, Sylvain, Tristan +3 · 7 citations
Computer Science · Medicine · #COVID-19 diagnosis using AI #Computer Vision and Pattern Recognition (cs.CV) #Domain Adaptation and Few-Shot Learning #FOS: Computer and information sciences #Machine Learning (cs.LG) #Multimodal Machine Learning Applications #cs.CV #cs.LG

paper · pdf · doi:10.48550/arxiv.1912.12179

Published at ICLR 2020

arxiv created 2019/12/20 · openalex publication_date 2019/12/20 · arxiv updated 2019/12/30 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

In this work we study locality and compositionality in the context of learning representations for Zero Shot Learning (ZSL). In order to well-isolate the importance of these properties in learned representations, we impose the additional constraint that, differently from most recent work in ZSL, no pre-training on different datasets (e.g. ImageNet) is performed. The results of our experiments show how locality, in terms of small parts of the input, and compositionality, i.e. how well can the learned representations be expressed as a function of a smaller vocabulary, are both deeply related to generalization and motivate the focus on more local-aware models in future research directions for representation learning.

Citations

• Auto-Encoding Variational Bayes
• Self-Attention Generative Adversarial Networks
• Visual Space Optimization for Zero-shot Learning
• Detecting Functionality-Specific Vulnerabilities via Retrieving Individual Functionality-Equivalent APIs in Open-Source Repositories

Cited by

• Learning Skill-Attributes for Transferable Assessment in Video
• Goal-Oriented Gaze Estimation for Zero-Shot Learning
• Simple and effective localized attribute representations for zero-shot learning
• Compositional Fine-Grained Low-Shot Learning
• Cross-Modal Information Maximization for Medical Imaging: CMIM
• Zero-Shot Learning from scratch (ZFS): leveraging local compositional representations
• ResNetVLLM -- Multi-modal Vision LLM for the Video Understanding Task

Related

• Zero-Shot Learning from scratch (ZFS): leveraging local compositional representations
• Learning Compositional Representations for Effective Low-Shot Generalization
• Learning Invariant Visual Representations for Compositional Zero-Shot Learning
• A causal view of compositional zero-shot recognition
• CORL: Compositional Representation Learning for Few-Shot Classification