The human auditory system is vulnerable to damage from aging, noise exposure, and genetic factors, often leading to hearing loss that conventional aids only partially remedy. This work uses computational auditory models to emulate sound processing and individual hearing impairments, enabling personalized compensation. Deep neural network (DNN) emulators are developed to mimic these models across diverse input ranges and hearing loss types. Using these emulators, novel DNN-based strategies for hearing-loss compensation and speech enhancement are proposed and validated through listening tests with hearing-impaired users.