Post-Hoc Domain Adaptation via Guided Data Homogenization
This addresses the challenge of domain adaptation for deployed models where parameter adjustments are not feasible, but it appears incremental as it builds on existing transfer learning concepts.
The paper tackles the problem of adapting deep learning models to new data distributions without modifying the model parameters, which is crucial for maintaining safety certifications in real-world deployments. They propose guided data homogenization, shifting adaptation from the model to the data, and demonstrate its potential on CIFAR-10 and MNIST datasets, though no concrete performance numbers are provided.
Addressing shifts in data distributions is an important prerequisite for the deployment of deep learning models to real-world settings. A general approach to this problem involves the adjustment of models to a new domain through transfer learning. However, in many cases, this is not applicable in a post-hoc manner to deployed models and further parameter adjustments jeopardize safety certifications that were established beforehand. In such a context, we propose to deal with changes in the data distribution via guided data homogenization which shifts the burden of adaptation from the model to the data. This approach makes use of information about the training data contained implicitly in the deep learning model to learn a domain transfer function. This allows for a targeted deployment of models to unknown scenarios without changing the model itself. We demonstrate the potential of data homogenization through experiments on the CIFAR-10 and MNIST data sets.