A new beat SQI based majority voting fusion method has been proposed for robust heart rate estimation from fusion of cardiovascular and non-cardiovascular (NC) signals. A novel statistical and probability based beat SQI assessment method has been proposed. It has achieved accuracy of 0.91 on PhysioNet/CinC Challenge-2011 set-a database. We have used Slope Sum Function and Teager-Kaiser Energy operator method for R-peak artifacts detection in NC signals. The majority voting fusion method has achieved score of 90.89% in heart beat detection on PhysioNet/CinC Challenge-2014 test dataset and 99.77% on MIT-BIH Polysomnographic dataset. The proposed method has substantially improved average HR rMSE from 15.54 bpm to 0.24 bpm for noisy ECG signals and from 11.68 bpm to 0.84 bpm for noisy ECG and noisy ABP signals of PhysioNet/CinC Challenge-2014 training database.