The ability to decode speech from brain activity has significant implications for brain-computer interfaces (BCIs), assistive communication technologies, and neuroscience research. EEG signals, which capture electrical activity in the brain, provide a non-invasive and portable method for studying neural responses to speech. However, decoding speech from EEG is challenging due to the low signal-to-noise ratio, the complex nature of speech processing in the brain, and the high dimensionality of EEG data. The acceptance student must work on a new method using a deep learning framework to decode the speech envelope (or speech content, depending on his/her focus) from EEG signals.