Monty A. Escabı́

The paper predicts speech recognition accuracy across diverse voices, words, and background noises using brain-inspired auditory features.

The paper investigates how background noise statistics affect speech encoding in the auditory midbrain of rabbits, finding that neural response variability depends on background spectrum and modulation statistics.

The paper develops parametric models to predict nonstationary spike-spike correlations using LFP features and applies them to track changes in correlations over time in mouse hippocampus and visual cortex data.

The paper shows that distinct neural populations in the auditory cortex encode task-related information differently, with nSTRF neurons contributing more to multisensory behavior.

The paper shows human speech recognition in natural noise relies on mid-level auditory statistics, with a model predicting perceptual variance better than cochlear models.

The paper investigates how bandwidth scaling in auditory filters improves the neural coding of natural sounds by whitening representations across frequency, time, and modulation dimensions.

The paper shows that distinct neural response statistics in the auditory system encode sound texture features for recognition and discrimination, with fast spectrum-based discrimination and slow correlation-based recognition.

The paper develops an artifact removal algorithm for neural recordings, improving signal-to-noise ratio by 20-40 dB across multiple stimulation scenarios.

The paper shows that correlation statistics between frequency-organized auditory channels can be used to identify sound categories with high accuracy, mirroring human performance.

A hierarchical spiking neural network achieves near human-level noise robust word recognition and mirrors properties of the auditory pathway.