UAV-IndoorCL

This repository contains the dataset presented in the paper Learning on the Fly: Replay-Based Continual Object Perception for Indoor Drones.

Project Page | GitHub

Dataset Summary

UAV-IndoorCL is an indoor video dataset consisting of 14,400 frames capturing inter-drone and ground vehicle footage. It was specifically designed to support and benchmark Class-Incremental Learning (CIL) research for resource-constrained aerial platforms. The frames were annotated via a semi-automatic workflow with high labeling agreement and final manual verification, ensuring temporal coherence across sequences.

Citation

If you use this dataset in your research, please cite the following paper:

@article{nae2026learningflyreplaybasedcontinual,
    title   = {Learning on the Fly: Replay-Based Continual Object Perception for Indoor Drones},
    author  = {Nae, Sebastian-Ion and Barbu, Mihai-Eugen and Mocanu, Sebastian and Leordeanu, Marius},
    journal = {arXiv preprint arXiv:2602.13440},
    year    = {2026}
}