Zero-Shot Learning via Class-Conditioned Deep Generative Models

Abstract

We present a deep generative model for learning to predict classes not seen at training time. Unlike most existing methods for this problem, that represent each class as a point (via a semantic embedding), we represent each seen/unseen class using a class-specific latent-space distribution, conditioned on class attributes. We use these latent-space distributions as a prior for a supervised variational autoencoder (VAE), which also facilitates learning highly discriminative feature representations for the inputs. The entire framework is learned end-to-end using only the seen-class training data. The model infers corresponding attributes of a test image by maximizing the VAE lower bound; the inferred attributes may be linked to labels not seen when training. We further extend our model to a (1) semi-supervised/transductive setting by leveraging unlabeled unseen-class data via an unsupervised learning module, and (2) few-shot learning where we also have a small number of labeled inputs from the unseen classes. We compare our model with several state-of-the-art methods through a comprehensive set of experiments on a variety of benchmark data sets.

To appear in AAAI 2018

References in corpus (13)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Stochastic Backpropagation and Approximate Inference in Deep Generative Models
• Matching Networks for One Shot Learning
• Zero-Shot Learning by Convex Combination of Semantic Embeddings
• Transductive Multi-view Zero-Shot Learning
• Semantic Autoencoder for Zero-Shot Learning
• Zero-Shot Learning via Semantic Similarity Embedding
• Zero-Shot Learning posed as a Missing Data Problem
• Learning Robust Visual-Semantic Embeddings
• Learning Joint Feature Adaptation for Zero-Shot Recognition
• Zero-Shot Learning with Generative Latent Prototype Model
• Zero-Shot Visual Recognition via Bidirectional Latent Embedding
• Transductive Zero-Shot Learning with a Self-training dictionary approach

Cited by in corpus (11)

• Transductive Zero-Shot Learning with Visual Structure Constraint
• Stacked Semantic-Guided Attention Model for Fine-Grained Zero-Shot Learning
• A Meta-Learning Framework for Generalized Zero-Shot Learning
• Zero and Few Shot Learning with Semantic Feature Synthesis and Competitive Learning
• Zero Shot Learning for Code Education: Rubric Sampling with Deep Learning Inference
• CRL: Class Representative Learning for Image Classification
• Graph-Driven Generative Models for Heterogeneous Multi-Task Learning
• Class Knowledge Overlay to Visual Feature Learning for Zero-Shot Image Classification
• Variationally Inferred Sampling Through a Refined Bound for Probabilistic Programs
• Zero-Shot Recognition via Optimal Transport
• Learning Graph-Based Priors for Generalized Zero-Shot Learning