Anomaly Detection for Autonomous Drones

Self-supervised multimodal anomaly detection using deep reconstruction and forecasting (IEEE SPC 2020 runner-up)

Self-supervised deep reconstruction and forecasting for anomaly detection in autonomous drone systems using multimodal sensor data (video + IMU).

  • Secured 2nd place globally in IEEE Signal Processing Cup 2020 (team Icarus Inhibition).
  • Achieved the highest score in the selection round.
  • Final competition held at ICASSP 2020.

Paper: Anomaly Detection using Deep Reconstruction and Forecasting for Autonomous Systems