Adaptive calibration for sensor-equipped athletic garments

Abstract

An exercise feedback system calibrates sensors of an athletic garment worn by an athlete while performing exercises. The sensors can record physiological data such as muscle activation. The system instructs the athlete to perform a calibration workout. The system generates a calibration value based on physiological data from the calibration workout and/or user information. The calibration value indicates, for example, the predicted maximum amplitude for the muscle activation of a particular muscle group (for example, glutes, hamstrings, or quadriceps) of the athlete. The system can update the calibration value over time as the system receives additional physiological data from subsequent exercises performed by the athlete. The system may determine a confidence level of the calibration value and may update the calibration value if the confidence level becomes too low. The system provides biofeedback to the athlete generated based on the calibration value.

What is claimed is: 
     
         1 . A method comprising:
 receiving a first set of physiological data from a garment worn by a user, the first set of physiological data describing muscle activation of a plurality of muscles while performing a workout, the garment including a plurality of sensors configured to generate physiological data;   receiving a second set of physiological data from the garment worn by the user, the second set of physiological data describing muscle activation of the plurality of muscles while performing a subsequent workout;   automatically refining a calibration value determined from the first set of physiological data, upon processing the second set of physiological data associated with the subsequent workout, wherein automatically refining the calibration value comprises:
 identifying peaks in the second set of physiological data; 
 for each peak, determining whether the peak is a first type or a second type; and 
 modifying the calibration value based on one or more peaks of the first type but not one or more peaks of the second type; and 
   providing biofeedback generated based on the refined calibration value to a client device of the user.   
     
     
         2 . The method of  claim 1 , wherein the one or more peaks of the first type are associated with a strength controlled exercise including multiple repetitions of an exercise. 
     
     
         3 . The method of  claim 1 , wherein the one or more peaks of the second type are associated with a power exercise including a single repetition of an exercise. 
     
     
         4 . The method of  claim 1 , wherein at least one peak of the one or more peaks of the second type is a transient peak. 
     
     
         5 . The method of  claim 1 , wherein the calibration value is determined based on a peak amplitude value in the first set of physiological data. 
     
     
         6 . The method of  claim 1 , wherein the calibration value is further determined based on at least one of user information, geographical location data of the user, a perceived effort level of the user, a performance metric of the user, and characteristics of the first calibration workout. 
     
     
         7 . The method of  claim 1 , wherein the refined calibration value is determined based on a peak amplitude value of the one or more peaks of the first type in the second set of physiological data. 
     
     
         8 . The method of  claim 7 , wherein the user information is received from a third party application associated with the client device of the user. 
     
     
         9 . The method of  claim 8 , wherein the first set of physiological data and at least one of the user information, geographical location data of the user, perceived effort level of the user, performance metric of the user, and characteristics of the workout are provided as input to a machine learning model that outputs the calibration value. 
     
     
         10 . The method of  claim 9 , wherein the machine learning model is trained using data collected from a population of athletes including physiological data associated with different perceived effort levels and target calibration values. 
     
     
         11 . The method of  claim 1 , further comprising:
 receiving a third set of physiological data from the garment;   determining a confidence level of the refined calibration value determined based on the second set of physiological data;   determining a difference between an amplitude of a data subset of the third set of physiological data and the refined calibration value; and   modifying the confidence level based on the difference of the amplitude of the data subset and the refined calibration value.   
     
     
         12 . The method of  claim 11 , wherein the confidence level is modified based on a machine learning model that generates a confidence level value based on the difference of the amplitude of the data subset and the refined calibration value, the machine learning model trained using feature vectors based on physiological data associated with a population of users that perform exercises, and modifying the confidence level includes reducing the confidence level by the confidence level value. 
     
     
         13 . A non-transitory computer-readable storage medium containing computer program code that, when executed by a processor, causes the processor to perform steps comprising:
 receiving a first set of physiological data from a garment worn by a user, the first set of physiological data describing muscle activation of a plurality of muscles while performing a workout, the garment including a plurality of sensors configured to generate physiological data;   receiving a second set of physiological data from the garment worn by the user, the second set of physiological data describing muscle activation of the plurality of muscles while performing a subsequent workout;   automatically refining a calibration value determined from the first set of physiological data, upon processing the second set of physiological data associated with the subsequent workout, wherein automatically refining the calibration value comprises:
 identifying peaks in the second set of physiological data; 
 for each peak, determining whether the peak is a first type or a second type; and 
 modifying the calibration value based on one or more peaks of the first type but not one or more peaks of the second type; and 
   providing biofeedback generated based on the refined calibration value to a client device of the user.   
     
     
         14 . The non-transitory computer-readable storage medium of  claim 13 , wherein the one or more peaks of the first type are associated with a strength controlled exercise including multiple repetitions of an exercise. 
     
     
         15 . The non-transitory computer-readable storage medium of  claim 13 , wherein the one or more peaks of the second type are associated with a power exercise including a single repetition of an exercise. 
     
     
         16 . The non-transitory computer-readable storage medium of  claim 13 , wherein at least one peak of the one or more peaks of the second type is a transient peak. 
     
     
         17 . The non-transitory computer-readable storage medium of  claim 13 , wherein the calibration value is determined based on a peak amplitude value in the first set of physiological data. 
     
     
         18 . The non-transitory computer-readable storage medium of  claim 13 , wherein the calibration value is further determined based on at least one of user information, geographical location data of the user, a perceived effort level of the user, a performance metric of the user, and characteristics of the first calibration workout. 
     
     
         19 . The non-transitory computer-readable storage medium of  claim 13 , wherein the refined calibration value is determined based on a peak amplitude value of the one or more peaks of the first type in the second set of physiological data. 
     
     
         20 . A system comprising:
 a processor; and   a non-transitory computer-readable storage medium containing computer program code that, when executed by a processor, causes the processor to perform steps comprising:
 receiving a first set of physiological data from a garment worn by a user, the first set of physiological data describing muscle activation of a plurality of muscles while performing a workout, the garment including a plurality of sensors configured to generate physiological data; 
 receiving a second set of physiological data from the garment worn by the user, the second set of physiological data describing muscle activation of the plurality of muscles while performing a subsequent workout; 
 automatically refining a calibration value determined from the first set of physiological data, upon processing the second set of physiological data associated with the subsequent workout, wherein automatically refining the calibration value comprises:
 identifying peaks in the second set of physiological data; 
 for each peak, determining whether the peak is a first type or a second type; and 
 modifying the calibration value based on one or more peaks of the first type but not one or more peaks of the second type; and 
 
 providing biofeedback generated based on the refined calibration value to a client device of the user.