A biometric information analysis method is provided. The biometric information analysis method comprises: an information reception step for receiving, by a server, sensing information; an exercise information generation step for generating, by the server, exercise information on the basis of the sensing information; an indicator calculation step for calculating, by the server, a performance indicator on the basis of the sensing information or the exercise information; a storage step for storing, by the server, the sensing information, the exercise information, or the performance indicator; a step for receiving, by the server, authentication information of a user input from a sensing device; an output information generation step for generating, by the server, output information on the basis of the sensing information, the exercise information, or the performance indicator; and an output information transmission step for transmitting, by the server, the output information to the sensing device or a client device, wherein the sensing information is acquired by sensing biometric information comprising respiration, electrocardiogram, oxygen saturation, body temperature, location, or motion information, the exercise information comprises information about an exercise type, the number of exercises, or exercise intensity, and the exercise information generation step or the indicator calculation step uses a deep learning algorithm.L'invention concerne un procédé d'analyse d'informations biométriques. Le procédé d'analyse d'informations biométriques comprend : une étape de réception d'informations pour recevoir, par un serveur, des informations de détection ; une étape de génération d'informations d'exercice pour générer, par le serveur, des informations d'exercice sur la base des informations de détection ; une étape de calcul d'indicateur pour calculer, par le serveur, un indicateur de performance sur la base des informations de détection ou des informations d'exercice ; un