A Note on Data Biases in Generative Models
This work is significant for researchers and practitioners in machine learning, particularly those working with generative models, by raising awareness and providing a tool to analyze and understand how data biases influence model outputs.
This paper investigates data biases in generative models by using a conditional invertible neural network to disentangle dataset-specific information from shared information. This allows projecting images across datasets to reveal inherent biases, showing how dataset quality impacts generative model performance and how societal biases are replicated.
It is tempting to think that machines are less prone to unfairness and prejudice. However, machine learning approaches compute their outputs based on data. While biases can enter at any stage of the development pipeline, models are particularly receptive to mirror biases of the datasets they are trained on and therefore do not necessarily reflect truths about the world but, primarily, truths about the data. To raise awareness about the relationship between modern algorithms and the data that shape them, we use a conditional invertible neural network to disentangle the dataset-specific information from the information which is shared across different datasets. In this way, we can project the same image onto different datasets, thereby revealing their inherent biases. We use this methodology to (i) investigate the impact of dataset quality on the performance of generative models, (ii) show how societal biases of datasets are replicated by generative models, and (iii) present creative applications through unpaired transfer between diverse datasets such as photographs, oil portraits, and animes. Our code and an interactive demonstration are available at https://github.com/CompVis/net2net .