Inverting face embeddings with convolutional neural networks

TL;DR

This paper presents methods to invert face embeddings into realistic images using neural networks, including gradient ascent with guidance and a trained network for real-time image generation.

Contribution

It introduces a neural network approach for inverting face embeddings, enabling real-time image synthesis and improved inversion quality.

Findings

Gradient ascent with guiding images can produce consistent face images.

A trained neural network can invert embeddings in a single pass for real-time results.

The neural network approach reduces inversion loss compared to gradient descent.

Abstract

Deep neural networks have dramatically advanced the state of the art for many areas of machine learning. Recently they have been shown to have a remarkable ability to generate highly complex visual artifacts such as images and text rather than simply recognize them. In this work we use neural networks to effectively invert low-dimensional face embeddings while producing realistically looking consistent images. Our contribution is twofold, first we show that a gradient ascent style approaches can be used to reproduce consistent images, with a help of a guiding image. Second, we demonstrate that we can train a separate neural network to effectively solve the minimization problem in one pass, and generate images in real-time. We then evaluate the loss imposed by using a neural network instead of the gradient descent by comparing the final values of the minimized loss function.