Systems and methods are disclosed for speech detection from facial skin movements. In one implementation, a system may include at least one light source, at least one sensor configured to receive light reflections from the at least one light source; and a processor configured to control the at least one light source to illuminate a region of a face of a user in a time interval. The processor may receive reflection signals indicative of light reflected from the face in the time interval. The reflection signals may be analyzed to determine facial skin movements in the time interval, wherein determining the sequence of words includes using an artificial neural network and a motion-to-language analysis. Then, based on the determined facial skin movements, the processor may determine a sequence of words associated with the facial skin movements, and output a textual transcription corresponding with the determined sequence of words.