Domain Generalization for Object Recognition with Multi-task Autoencoders

Abstract

The problem of domain generalization is to take knowledge acquired from a number of related domains where training data is available, and to then successfully apply it to previously unseen domains. We propose a new feature learning algorithm, Multi-Task Autoencoder (MTAE), that provides good generalization performance for cross-domain object recognition. Our algorithm extends the standard denoising autoencoder framework by substituting artificially induced corruption with naturally occurring inter-domain variability in the appearance of objects. Instead of reconstructing images from noisy versions, MTAE learns to transform the original image into analogs in multiple related domains. It thereby learns features that are robust to variations across domains. The learnt features are then used as inputs to a classifier. We evaluated the performance of the algorithm on benchmark image recognition datasets, where the task is to learn features from multiple datasets and to then predict the image label from unseen datasets. We found that (denoising) MTAE outperforms alternative autoencoder-based models as well as the current state-of-the-art algorithms for domain generalization.

accepted in ICCV 2015

References in corpus (1)

• Sequence to Sequence Learning with Neural Networks

Cited by in corpus (31)

• Deep Visual Domain Adaptation: A Survey
• Generalizing Across Domains via Cross-Gradient Training
• Domain Adaptation for Visual Applications: A Comprehensive Survey
• Domain Generalization by Marginal Transfer Learning
• Learning to Generalize: Meta-Learning for Domain Generalization
• Domain Generalization via Optimal Transport with Metric Similarity Learning
• Sample-to-Sample Correspondence for Unsupervised Domain Adaptation
• Deep Cocktail Network: Multi-source Unsupervised Domain Adaptation with Category Shift
• Domain Generalization Using a Mixture of Multiple Latent Domains
• Scatter Component Analysis: A Unified Framework for Domain Adaptation and Domain Generalization
• Zero-shot Domain Adaptation without Domain Semantic Descriptors
• Variational Disentanglement for Domain Generalization
• Dual Reweighting Domain Generalization for Face Presentation Attack Detection
• Supervised COSMOS Autoencoder: Learning Beyond the Euclidean Loss!
• Unsupervised Domain Adaptation: from Simulation Engine to the RealWorld
• Structure-Attentioned Memory Network for Monocular Depth Estimation
• Domain2Vec: Deep Domain Generalization
• Deeper, Broader and Artier Domain Generalization
• Dynamic Feature Alignment for Semi-supervised Domain Adaptation
• Unsupervised Domain Adaptation using Generative Models and Self-ensembling
• Robustness through Data Augmentation Loss Consistency
• Multi-source Domain Adaptation for Visual Sentiment Classification
• Efficient Action Detection in Untrimmed Videos via Multi-Task Learning
• EEG-Based Driver Drowsiness Estimation Using Feature Weighted Episodic Training
• Embracing the Dark Knowledge: Domain Generalization Using Regularized Knowledge Distillation
• Zero Shot Domain Generalization
• Semi-Supervised Adversarial Discriminative Domain Adaptation
• Deep Shape Matching
• Action-Affect Classification and Morphing using Multi-Task Representation Learning
• KL Guided Domain Adaptation
• Multi-component Image Translation for Deep Domain Generalization