Content Bias in Deep Learning Image Age Approximation: A new Approach Towards better Explainability
This addresses the issue of explainability and reliability in temporal image forensics for researchers and practitioners, but it is incremental as it builds on existing methods to improve bias detection.
The paper tackles the problem of content bias in deep learning models for image age approximation, showing that existing methods likely rely heavily on image content rather than age-related features, as verified using synthetic images. It proposes evaluating content influence and tests countermeasures like steganalysis models and preprocessing techniques to enhance the age signal.
In the context of temporal image forensics, it is not evident that a neural network, trained on images from different time-slots (classes), exploits solely image age related features. Usually, images taken in close temporal proximity (e.g., belonging to the same age class) share some common content properties. Such content bias can be exploited by a neural network. In this work, a novel approach is proposed that evaluates the influence of image content. This approach is verified using synthetic images (where content bias can be ruled out) with an age signal embedded. Based on the proposed approach, it is shown that a deep learning approach proposed in the context of age classification is most likely highly dependent on the image content. As a possible countermeasure, two different models from the field of image steganalysis, along with three different preprocessing techniques to increase the signal-to-noise ratio (age signal to image content), are evaluated using the proposed method.