Heart rate monitors are plagued by noisy photoplethysmography (PPG) data, which makes it difficult for the monitors to output a consistently accurate heart rate reading. Noise is often caused by motion. Using known methods for processing accelerometer readings that measure movement to filter out some of this noise may help, but not always. The present disclosure describes an improved front-end technique (time-domain interference removal) based on using adaptive linear prediction on accelerometer data to generate filters for filtering the PPG signal prior to tracking the frequency of the heartbeat (heart rate). The present disclosure also describes an improved back-end technique based on steering the frequency of a resonant filter in order to track the heartbeat. Implementing one or both of these techniques leads to more accurate heart rate measurements.