MOSAD: Multimodal Open-Set Anomaly Detection Dataset for Omnidirectional UAVs

This repository contains the flight sequences used in the MOSAD (Multimodal Open-Set Anomaly Detection) framework, as described in the paper "Multimodal Open-Set Anomaly Detection for Omnidirectional UAVs via PCA Gating and Bayesian Latent Regularization". Dataset Details: Total sequences: 8 flight sequences (recorded from an omnidirectional hexacopter). Data format: CSV. Content: Includes nominal flights, visual anomalies, and mechanical degradations (5%-15% single-motor thrust loss). Sampling: Frames resized to 64x64 pixels. Please cite the associated paper if you use this dataset in your researc