Collaborative Visual Place Recognition through Federated Learning
This work addresses the problem of decentralized training for VPR, which is incremental as it adapts existing methods to a federated setting.
The paper tackles the problem of adapting Visual Place Recognition (VPR) to Federated Learning (FL) by proposing the FedVPR framework, which addresses challenges like lack of well-defined classes and client heterogeneity, resulting in a new task for FL research.
Visual Place Recognition (VPR) aims to estimate the location of an image by treating it as a retrieval problem. VPR uses a database of geo-tagged images and leverages deep neural networks to extract a global representation, called descriptor, from each image. While the training data for VPR models often originates from diverse, geographically scattered sources (geo-tagged images), the training process itself is typically assumed to be centralized. This research revisits the task of VPR through the lens of Federated Learning (FL), addressing several key challenges associated with this adaptation. VPR data inherently lacks well-defined classes, and models are typically trained using contrastive learning, which necessitates a data mining step on a centralized database. Additionally, client devices in federated systems can be highly heterogeneous in terms of their processing capabilities. The proposed FedVPR framework not only presents a novel approach for VPR but also introduces a new, challenging, and realistic task for FL research, paving the way to other image retrieval tasks in FL.