Two-level Data Augmentation for Calibrated Multi-view Detection
This work addresses the challenge of scarce and expensive-to-annotate multi-view data for computer vision applications, offering a practical solution for improved detection.
The paper tackles the problem of data augmentation breaking alignment in multi-view systems by introducing a two-level augmentation pipeline that preserves view alignment, resulting in significant performance improvements over existing baselines on WILDTRACK and MultiviewX datasets.
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.