Context-adaptable radar-based people counting via few-shot learning
Metadatos
Afficher la notice complèteEditorial
Springer Nature
Materia
Active learning Meta learning Radar Few shot learning People counting Weighting network
Date
2023-08-08Referencia bibliográfica
Mauro, G., Martinez-Rodriguez, I., Ott, J. et al. Context-adaptable radar-based people counting via few-shot learning. Appl Intell (2023). [https://doi.org/10.1007/s10489-023-04778-z]
Patrocinador
ECSEL Joint Under-taking (JU) 876925; Horizon 2020; Universidad de Granada/CBUARésumé
In many industrial or healthcare contexts, keeping track of the number of people is essential. Radar systems, with their low overall cost and power consumption, enable privacy-friendly monitoring in many use cases. Yet, radar data are hard to interpret and incompatible with most computer vision strategies. Many current deep learning-based systems achieve high monitoring performance but are strongly context-dependent. In this work, we show how context generalization approaches can let the monitoring system fit unseen radar scenarios without adaptation steps. We collect data via a 60 GHz frequency-modulated continuous wave in three office rooms with up to three people and preprocess them in the frequency domain. Then, using meta learning, specifically the Weighting-Injection Net, we generate relationship scores between the few training datasets and query data. We further present an optimization-based approach coupled with weighting networks that can increase the training stability when only very few training examples are available. Finally, we use pool-based sampling active learning to fine-tune the model in new scenarios, labeling only the most uncertain data. Without adaptation needs, we achieve over 80% and 70% accuracy by testing the meta learning algorithms in new radar positions and a new office, respectively.