Low Complex IIR Adaptive Hear-Through Ambient Filtering for Overcoming Practical Constraints in Earbuds

The materials used in the headphones and earbuds typically attenuate sound passively above frequency of 500-700 Hz. Hear-Through (HT) techniques can compensate for the passive attenuation characteristics of listening devices to achieve a listening experience similar to the open ear for enhancing situational awareness of the users. Most studies have proposed usage of fixed-filter or adaptive finite impulse response (FIR) HT techniques, which can lead to mismatch between the modelled and desired open ear response for different users and fittings or lead to higher computations and larger processing delays. This paper proposes a virtual sensing model with an infinite impulse response (IIR) based adaptive equalization method to model the open ear response at user’s eardrum. The proposed method reduces the computational complexity by 34%, while maintaining stability and providing superior HT performance upto 10 dB compared to existing methods in dynamic acoustic scenarios.