# Looking Beyond Appearances: Synthetic Training Data for Deep CNNs in Re-identification

## TL;DR

This paper introduces SOMAnet, a deep CNN trained on a large synthetic dataset, SOMAset, to improve person re-identification by modeling structural attributes beyond clothing, achieving superior performance.

## Contribution

It presents a novel Inception-based CNN architecture trained on a synthetic dataset, enabling re-identification across clothing variations and reducing data preparation complexity.

## Key findings

- SOMAnet outperforms existing methods on re-identification benchmarks.
- Synthetic SOMAset data effectively trains the model, reducing reliance on real paired images.
- The approach enables re-identification despite clothing changes.

## Abstract

Re-identification is generally carried out by encoding the appearance of a subject in terms of outfit, suggesting scenarios where people do not change their attire. In this paper we overcome this restriction, by proposing a framework based on a deep convolutional neural network, SOMAnet, that additionally models other discriminative aspects, namely, structural attributes of the human figure (e.g. height, obesity, gender). Our method is unique in many respects. First, SOMAnet is based on the Inception architecture, departing from the usual siamese framework. This spares expensive data preparation (pairing images across cameras) and allows the understanding of what the network learned. Second, and most notably, the training data consists of a synthetic 100K instance dataset, SOMAset, created by photorealistic human body generation software. Synthetic data represents a good compromise between realistic imagery, usually not required in re-identification since surveillance cameras capture low-resolution silhouettes, and complete control of the samples, which is useful in order to customize the data w.r.t. the surveillance scenario at-hand, e.g. ethnicity. SOMAnet, trained on SOMAset and fine-tuned on recent re-identification benchmarks, outperforms all competitors, matching subjects even with different apparel. The combination of synthetic data with Inception architectures opens up new research avenues in re-identification.