Learning Classifiers for Domain Adaptation, Zero and Few-Shot Recognition Based on Learning Latent Semantic Parts
arXiv:1901.09079
Abstract
In computer vision applications, such as domain adaptation (DA), few shot learning (FSL) and zero-shot learning (ZSL), we encounter new objects and environments, for which insufficient examples exist to allow for training "models from scratch," and methods that adapt existing models, trained on the presented training environment, to the new scenario are required. We propose a novel visual attribute encoding method that encodes each image as a low-dimensional probability vector composed of prototypical part-type probabilities. The prototypes are learnt to be representative of all training data. At test-time we utilize this encoding as an input to a classifier. At test-time we freeze the encoder and only learn/adapt the classifier component to limited annotated labels in FSL; new semantic attributes in ZSL. We conduct extensive experiments on benchmark datasets. Our method outperforms state-of-art methods trained for the specific contexts (ZSL, FSL, DA).
arXiv comment9 pages, 3 figures, 3 tables. Accepted at ICML 2019
References in corpus (23)
- Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
- Prototypical Networks for Few-shot Learning
- Conditional Adversarial Domain Adaptation
- Label-Embedding for Image Classification
- Data Augmentation Generative Adversarial Networks
- Asymmetric Tri-training for Unsupervised Domain Adaptation
- A DIRT-T Approach to Unsupervised Domain Adaptation
- Unsupervised Cross-Domain Image Generation
- Meta-Learning with Latent Embedding Optimization
- Semantic Autoencoder for Zero-Shot Learning
- Co-regularized Alignment for Unsupervised Domain Adaptation
- Few-Shot Learning via Embedding Adaptation with Set-to-Set Functions
- Generative Adversarial Residual Pairwise Networks for One Shot Learning
- Preserving Semantic Relations for Zero-Shot Learning
- An Empirical Study and Analysis of Generalized Zero-Shot Learning for Object Recognition in the Wild
- Low-Shot Learning from Imaginary Data
- Self-ensembling for visual domain adaptation
- Domain-Invariant Projection Learning for Zero-Shot Recognition
- Multi-modal Ensemble Classification for Generalized Zero Shot Learning
- Deep Adversarial Attention Alignment for Unsupervised Domain Adaptation: the Benefit of Target Expectation Maximization
- Adaptive Confidence Smoothing for Generalized Zero-Shot Learning
- Transfer of View-manifold Learning to Similarity Perception of Novel Objects
- Generative Dual Adversarial Network for Generalized Zero-shot Learning