Generative One-Class Models for Text-based Person Retrieval in Forensic Applications

TL;DR

This paper introduces a generative one-class model for text-based person retrieval in forensic images, emphasizing efficiency, robustness, and low training data requirements, and compares it to a discriminative approach.

Contribution

It presents a novel generative color model with outlier filtering for forensic person retrieval using textual descriptions, highlighting its advantages over discriminative methods.

Findings

Generative model performs comparably to discriminative models in retrieval accuracy.

The proposed method requires fewer training samples.

It has lower computational costs with increasing training data.

Abstract

Automatic forensic image analysis assists criminal investigation experts in the search for suspicious persons, abnormal behaviors detection and identity matching in images. In this paper we propose a person retrieval system that uses textual queries (e.g., "black trousers and green shirt") as descriptions and a one-class generative color model with outlier filtering to represent the images both to train the models and to perform the search. The method is evaluated in terms of its efficiency in fulfilling the needs of a forensic retrieval system: limited annotation, robustness, extensibility, adaptability and computational cost. The proposed generative method is compared to a corresponding discriminative approach. Experiments are carried out using a range of queries in three different databases. The experiments show that the two evaluated algorithms provide average retrieval performance and adaptable to new datasets. The proposed generative algorithm has some advantages over the discriminative one, specifically its capability to work with very few training samples and its much lower computational requirements when the number of training examples increases.