The ability to categorize and recognize sounds is an amazing trait of the mammalian brain. We can easily recognize a familiar voice, identify a musical tune, or segregate a single speaker in a crowded room. Yet these seemingly simple tasks present major challenges to our most sophisticated computers, which fail miserably at such sound recognition tasks especially in the presence of background noise. Our lab uses large-scale neural recordings and computational models to explore how the brain achieves such seemingly simple feats. By understanding the computational principles performed during natural hearing, we hope to develop superior sound recognition technologies and treatment strategies that reverse the detrimental effects of hearing loss.