A non-intrusive signal quality assessment apparatus includes a feature vector calculator that determines parameters representing frames of a signal and extracts a collection of per-frame feature vectors (φ;(n)) representing structural information of the signal from the parameters. A frame selector preferably selects only frames (Ω\with a feature vector (φ;(n)) lying within a predetermined multi-dimensional window (Θ). Means determine a global feature set (ψ) over the collection of feature vectors (φ;(n)) from statistical moments of selected feature vector components ((1^,02, . . . O11). A quality predictor predicts a signal quality measure (Qj from the global feature set (ψ)).