Two-level Data Augmentation for Calibrated Multi-view Detection

TL;DR

This paper introduces a novel two-level data augmentation pipeline for multi-view detection that maintains view alignment and significantly improves detection performance on key datasets.

Contribution

It presents a new multi-view data augmentation method that preserves view alignment and enhances detection accuracy, addressing a key challenge in multi-view systems.

Findings

Outperforms existing baselines on WILDTRACK and MultiviewX datasets

Effective preservation of view alignment during augmentation

Significant performance gains with simple multi-view detection model

Abstract

Data augmentation has proven its usefulness to improve model generalization and performance. While it is commonly applied in computer vision application when it comes to multi-view systems, it is rarely used. Indeed geometric data augmentation can break the alignment among views. This is problematic since multi-view data tend to be scarce and it is expensive to annotate. In this work we propose to solve this issue by introducing a new multi-view data augmentation pipeline that preserves alignment among views. Additionally to traditional augmentation of the input image we also propose a second level of augmentation applied directly at the scene level. When combined with our simple multi-view detection model, our two-level augmentation pipeline outperforms all existing baselines by a significant margin on the two main multi-view multi-person detection datasets WILDTRACK and MultiviewX.