Adapting Convolutional Neural Networks for Geographical Domain Shift
This addresses domain shift for image classification practitioners, offering a simple, incremental solution for scenarios with limited target data.
The paper tackled geographical domain shift in image processing by fine-tuning the last layer of an ensemble of CNNs using a small labeled target dataset, achieving first place in the NeurIPS 2018 Inclusive Images Competition.
We present the winning solution for the Inclusive Images Competition organized as part of the Conference on Neural Information Processing Systems (NeurIPS 2018) Competition Track. The competition was organized to study ways to cope with domain shift in image processing, specifically geographical shift: the training and two test sets in the competition had different geographical distributions. Our solution has proven to be relatively straightforward and simple: it is an ensemble of several CNNs where only the last layer is fine-tuned with the help of a small labeled set of tuning labels made available by the organizers. We believe that while domain shift remains a formidable problem, our approach opens up new possibilities for alleviating this problem in practice, where small labeled datasets from the target domain are usually either available or can be obtained and labeled cheaply.