Fig. 5: Block diagrams of popular SELD model architectures.
From: Environmental acoustic intelligence through sound event localization and detection: a review
(Left) The SELDNet architecture, adapted from Adavanne et al.5 It consists of convolutional layers, biGRU recurrent modules, and fully-connected layers to produce frame-by-frame event activity probabilities and corresponding DOA Cartesian coordinates. (Right) Block diagram of the popular ResNet-Conformer hybrid architecture, adapted from ref. 74, used by many top-performing systems.
