No abstract

The ability to discriminate between original documents and their photocopies poses a problem when conducting automated forensic examinations of large numbers of confiscated documents. This paper describes a novel frequency domain approach for printing technique and copy detection of scanned document images. Tests using a dataset consisting of 49 laser-printed, 14 inkjet-printed and 46 photocopied documents demonstrate that the approach outperforms existing spatial domain methods for image resolutions exceeding 200 dpi. An increase in classification accuracy of approximately 5% is achieved for low scan resolutions of 300 dpi and 400 dpi. In addition, the approach has the advantage of increased processing speed.

The detection of fraud in accounting data is a long-standing challenge in financial statement audits. Nowadays, the majority of applied techniques refer to handcrafted rules derived from known fraud scenarios. While fairly successful, these rules exhibit the drawback that they often fail to generalize beyond known fraud scenarios and fraudsters gradually find ways to circumvent them. In contrast, more advanced approaches inspired by the recent success of deep learning often lack seamless interpretability of the detected results. To overcome this challenge, we propose the application of adversarial autoencoder networks. We demonstrate that such artificial neural networks are capable of learning a semantic meaningful representation of real-world journal entries. The learned representation provides a holistic view on a given set of journal entries and significantly improves the interpretability of detected accounting anomalies. We show that such a representation combined with the networks reconstruction error can be utilized as an unsupervised and highly adaptive anomaly assessment. Experiments on two datasets and initial feedback received by forensic accountants underpinned the effectiveness of the approach.

No abstract