IIT-Madras Aims To Interpret Brain Signals Into Language

On Monday, the IIT-Madras announced that a team of researchers working there has geared up an AI technology in search of converting our brain signals of speech into language.

In general, brain signals are shaped in various waveforms, such as our responses to external forces or the plant of the photosynthesis process. In fact, any signs of signals from electrical signals to brain signals can be decoded and interpreted into meaningful information thanks to the use of mathematical transforms and physical laws such as Laplace Transform or Fourier Transform, which are named after famous mathematicians Jean-Baptiste Joseph Fourier and Pierre-Simon Laplace.

According to Vishal Nandigana, an Assistant Professor at IIT Madras, the processing stages will result in electrically driven ionic signals, which show the flow of ions being charged particles. From that, the team will have data to work on and interpret them into human language meaning speech.

Nandigana added that the output will be paired with a dedicated library of electrophysiological data to understand the brain signals from humans’ thinking, especially what they’re trying to deliver in speech. In the early stage, researchers focus on decoding real data signals and how they can link them with a simple human language thanks to a slew of advanced algorithms in Deep Learning and Artificial Intelligence.

Electrical signals from our brain appear to be wave-like patterns, so it’s evident that reading these signals is possible. In fact, researchers once tested several experiments with nanofluidic and nanopores, which also ended up with useful signals, according to a statement filed by the IIT-Madras.