System and Methods for Locally Customizing Media Content for Rendering

Systems, methods and devices process received media content to generate personalized media presentations on an end point device. Received media content may be buffered in a moving window buffer, and processed to create tokens by parsing a next content element, and, for each content element, identifying a speaker or actor, creating a text representation, and measuring perceptual properties such as pitch, timbre, volume, timing, and frame rate. The end point device may compare a segment of tokens within buffered media content to a list of replacement subject matter within a user profile to determine whether the segment matches any of the replacement subject matter, and identify substitute subject matter for the matched replacement subject matter. The end point device may create a replacement sequence by modifying the substitute subject matter using the perceptual properties of the tokens in the segment, and render a personalized media presentation including the replacement sequence.