The present disclosure relates to a pair of smart glasses. The pair of smart glasses may include a frame, one or more lenses, and one or more temples. The pair of smart glasses may further include at least one microphone array, a noise reduction device, at least one low-frequency acoustic driver, at least one high-frequency acoustic driver, and a controller. The at least one microphone array may be configured to generate sub-band voice signals. The noise reduction device may be configured to generate a noise correction signal based on each of the sub-band voice signal and generate a sub-band correction signal based on the noise correction signal and the sub-band voice signal. The controller may be configured to direct the low-frequency acoustic driver to output the sounds in a first frequency range and direct the high-frequency acoustic driver to output the sounds in a second frequency range.