# AutoDIAL: Automatic DomaIn Alignment Layers

## TL;DR

AutoDIAL introduces a novel method for domain adaptation by automatically learning where and how to align feature representations across domains within deep networks, improving performance on varied datasets.

## Contribution

It proposes Domain Alignment Layers that automatically determine the level of feature alignment needed at different network layers, unlike prior methods with fixed adaptation points.

## Key findings

- Effective in unsupervised domain adaptation benchmarks
- Automatically learns optimal feature alignment levels
- Outperforms traditional fixed-layer adaptation methods

## Abstract

Classifiers trained on given databases perform poorly when tested on data acquired in different settings. This is explained in domain adaptation through a shift among distributions of the source and target domains. Attempts to align them have traditionally resulted in works reducing the domain shift by introducing appropriate loss terms, measuring the discrepancies between source and target distributions, in the objective function. Here we take a different route, proposing to align the learned representations by embedding in any given network specific Domain Alignment Layers, designed to match the source and target feature distributions to a reference one. Opposite to previous works which define a priori in which layers adaptation should be performed, our method is able to automatically learn the degree of feature alignment required at different levels of the deep network. Thorough experiments on different public benchmarks, in the unsupervised setting, confirm the power of our approach.