Deep Feature Interpolation for Image Content Changes

TL;DR

Deep Feature Interpolation (DFI) is a simple, data-driven method that uses linear interpolation of deep features from pre-trained networks to perform high-resolution image transformations, often matching or surpassing more complex methods.

Contribution

DFI introduces a straightforward baseline for image transformation that requires no training of specialized networks, demonstrating high-level semantic edits with simple linear interpolation.

Findings

DFI can perform semantic transformations like aging, adding glasses, or smiling.

DFI sometimes outperforms state-of-the-art methods.

DFI serves as a practical baseline for evaluating image transformation tasks.

Abstract

We propose Deep Feature Interpolation (DFI), a new data-driven baseline for automatic high-resolution image transformation. As the name suggests, it relies only on simple linear interpolation of deep convolutional features from pre-trained convnets. We show that despite its simplicity, DFI can perform high-level semantic transformations like "make older/younger", "make bespectacled", "add smile", among others, surprisingly well - sometimes even matching or outperforming the state-of-the-art. This is particularly unexpected as DFI requires no specialized network architecture or even any deep network to be trained for these tasks. DFI therefore can be used as a new baseline to evaluate more complex algorithms and provides a practical answer to the question of which image transformation tasks are still challenging in the rise of deep learning.