File:Application of a bioinspired neuromorphic motion-cognition nerve in recognition of drone flight modes.webp

English: "a Motion-cognition quadrotor drone equipped with the IMU sensor, which wirelessly transmits bimodal motion signals to the spike-encoding circuit and synaptic transistor. Recognition of drone flight modes was implemented by spike coding and synaptic processing of the motion signals. b–f Postsynaptic current (PSC) generated from the synaptic transistor, when the drone was operated in five modes: b wind stabilization, c collision stabilization, d spinning, e ascending, and f spin-ascending. g Visualization of the two hardware outputs (mean firing rate of acceleration spikes facc, mean firing rate of gyroscopic spikes fgyro) in 2D coordinates. h Visualization of the three hardware outputs (facc, fgyro, and synaptic response ΔPSC) in 3D coordinates. The data clusters in different colors correspond to the five flight modes. i Confusion matrix showing the results of drone flight-mode recognition."