Energy-based and modulation-based models predict consonant recognition but not confusions.

Evidence

A few studies have attempted to simulate consonant perception. Li et al. ( 2010 ) successfully related consonant recognition data to the so-called AI Gram, which is related to the energy-based Articulation Index (ANSI 1969 ). Gallun and Souza ( 2008 ) considered noise-vocoded VCVs and demonstrated that the correlation of long-term modulation power representations was a strong predictor of consonant confusions. Jürgens and Brand ( 2009 ) used an auditory model with a modulation-frequency selective preprocessing stage in combination with a template-matching back end. The model showed convincing recognition predictions while the confusion predictions were inconclusive. — Dijk, Pim van, Physiology, Psychoacoustics and…, 5578

Categorization

• corroborated: 3 independent passages support this claim — Dijk, Pim van, Physiology, Psychoacoustics and…, 5578
• XREF: connects to: Subband and modulation power both strongly predict perceptual distances in consonant perception; Slow envelope fluctuations up to 16 Hz predict consonant-in-noise perception. — Dijk, Pim van, Physiology, Psychoacoustics and…, 5575; Consonant confusions shift with noise type, filtering, and vowel context. — Dijk, Pim van, Physiology, Psychoacoustics and…, 5577 (+1 more)

Broader Connections

• Subband and modulation power both strongly predict perceptual distances in consonant perception #corroborates (0.85)
• Consonant confusions shift with noise type, filtering, and vowel context. #expands (0.82) — Dijk, Pim van, Physiology, Psychoacoustics and…, 5577
• Slow envelope fluctuations up to 16 Hz predict consonant-in-noise perception. #corroborates (0.78) — Dijk, Pim van, Physiology, Psychoacoustics and…, 5575
• Speech intelligibility in rooms can be predicted from fluctuation strength #expands (0.75) — Professor Dr.-Ing. Hugo Fastl, Professor Dr.-Ing. Eberhard, Psychoacoustics- Facts and Models, 2059