Footwear outsole images predict wearer sex using convolutional neural networks
Sex Estimation from Footwear Outsole Impressions Using CNN Transfer Learning and Interpretable Image Statistics
Computer Vision and Pattern RecognitionMachine Learning
Summary
Footwear outsole impressions left at crime scenes can reveal information about the person who wore them, including their sex. The authors compare traditional methods with newer computer vision techniques using pretrained convolutional neural networks (CNNs). They find that fine-tuned CNNs better classify the sex of the wearer from shoe prints than older feature-based methods. The study also links the CNN’s internal features to measurable image properties, making the results more understandable. More testing on real casework images is needed before it can be used in practice.
What this means in practice
- •For forensic investigators: Support sex estimation of suspects from footwear impressions to improve forensic screening efficiency.
- •For security screening teams: Enhance biometric screening by incorporating automated sex prediction from shoeprints found in access control areas.
Authors
Jinyi Niu, Ziyi Song, Weining Shen
Abstract
Footwear outsole impressions are a common form of forensic pattern evidence, yet quantitative methods for estimating wearer attributes from these images remain relatively underdeveloped. We investigate binary sex estimation from footwear outsole impressions by comparing convolutional neural network (CNN) transfer learning with traditional feature-based classification. Using a publicly available outsole-impression dataset, we adopt a shoe-level training and test partition that keeps replicate scans of the same physical shoe together to reduce data leakage. We evaluate pretrained CNNs through end-to-end fine-tuning, frozen feature extraction followed by support vector machine classification, and hybrid feature fusion incorporating handcrafted, geometric, and metadata-derived descriptors. Fine-tuned CNNs achieve the strongest overall predictive performance and substantially outperform traditional classifiers trained on the manually specified descriptors alone, while frozen-feature approaches offer a less computationally demanding alternative. Exploratory analysis of low-dimensional CNN representations reveals associations with frequency threshold ratio, image contrast, and wavelet-based summaries, providing a connection between learned representations and measurable properties of outsole impressions. These findings suggest that CNN transfer learning captures discriminative information beyond the descriptors considered and offers a promising approach to footwear-based forensic screening. Further validation on independently collected and casework-like impressions is needed before operational use.