HeadFi technology, which is being developed by a team at New Jersey’s Rutgers University, will add the features of the most advanced smart models to simple headphones. The fact is that any headphones form a closed volume of air between the sound-emitting membrane on one side and the eardrum of their owner on the other. Moreover, the geometry and mechanical properties of this cavity are purely individual.
Information about this volume is encrypted in the pressure changes inside it. And they, in turn, make the headphone speaker generate the corresponding signal. This is what the HeadFi system receives and then effectively processes: it is an adapter module that is plugged in between the headphones and an audio source (for example, a smartphone).
Pressure fluctuations can be caused by external sounds, touching, and even a heartbeat (pulse). And the HeadFi system is ready to decipher any of these influences. The researchers tested several pairs of on-ear headphones ranging from $ 3 to $ 15,000, 54 sets, with 53 volunteers. It turned out that the system recognizes the headphone listener with an accuracy of 97.2% to 99.5%. She measured her heart rate with an accuracy of 96.8% to 99.2%, and hand touches – from 97.7% to 99.3%.
If the HeadFi turns out to be as efficient and cost effective in mass production, the system can add very interesting perspectives to the headphones. A more compact version of the prototype is currently in development. More detailed work on HeadFi from researchers at the University of Science and Technology of China, University of Massachusetts Amherst, Microsoft and Alibaba Group will be presented at MobiCom 2 2021 in October.