Basic vital signs such as heart and respiratory rates (HR and RR) are essential bio-indicators. Their longitudinal in-home collection enables prediction and detection of disease onset and change, providing for earlier health intervention. This type of data collection, interpretation and evaluation is especially valuable for older adults facing myriads of health challenges. However, respiration harmonics and intermodulation cause strong disturbances to much weaker heartbeat signals, thus robust vital signs monitoring remains elusive. In this paper, we propose VitalHub, a robust, non-touch vital signs monitoring system using a pair of co-located Ultra-Wide Band (UWB) and depth sensors. By extensive manual examination, we identify four typical temporal and spectral signal patterns and their suitable vital signs estimators. We devise a probabilistic weighted framework (PWF) that quantifies evidence of these patterns to update the weighted combination of estimator output to track the vital signs robustly. We also design a 'heatmap' based signal quality detector that achieves near-human performance differentiating signal corruptions from large motion. To monitor multiple cohabiting subjects in-home, we leverage consecutive skeletal poses from the depth data to distinguish between individuals and their activities, providing activity context important to disambiguating critical from normal vital sign variability. Extensive experiments show that VitalHub achieves 1.5/3.2 'breaths/beats per minute' (denoted by 'bpm') errors at 80-percentile for RR/HR, approaching the 1.2/1.5 bpm error 'ceiling' of an idealistic but impractical oracle. We also reveal how existing techniques for harmonics and intermodulation rely on presumed signal patterns thus may fail under real-world dynamic changes.