English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Roadside unit (RSU) can significantly improve the safety and robustness ofautonomous vehicles through Vehicle-to-Everything (V2X) communication.Currently, the usage of a single RSU mainly focuses on real-time inference andV2X collaboration, while neglecting the potential value of the high-qualitydata collected by RSU sensors. Integrating the vast amounts of data fromnumerous RSUs can provide a rich source of data for model training. However,the absence of ground truth annotations and the difficulty of transmittingenormous volumes of data are two inevitable barriers to fully exploiting thishidden value. In this paper, we introduce FedRSU, an innovative federatedlearning framework for self-supervised scene flow estimation. In FedRSU, wepresent a recurrent self-supervision training paradigm, where for each RSU, thescene flow prediction of points at every timestamp can be supervised by itssubsequent future multi-modality observation. Another key component of FedRSUis federated learning, where multiple devices collaboratively train an ML modelwhile keeping the training data local and private. With the power of therecurrent self-supervised learning paradigm, FL is able to leverage innumerableunderutilized data from RSU. To verify the FedRSU framework, we construct alarge-scale multi-modality dataset RSU-SF. The dataset consists of 17 RSUclients, covering various scenarios, modalities, and sensor settings. Based onRSU-SF, we show that FedRSU can greatly improve model performance in ITS andprovide a comprehensive benchmark under diverse FL scenarios. To the best ofour knowledge, we provide the first real-world LiDAR-camera multi-modal datasetand benchmark for the FL community.

FedRSU: Federated Learning for Scene Flow Estimation on Roadside Units