University of Patras

The EEG foundation model REVE no longer needs full fine-tuning or all-layer pooled features to recognize emotion. Such models had not consistently beaten task-specific architectures on cross-subject tasks, and their large representations are costly to compute and store. This approach aggregates the first-layer attention weights of a single head over each trial into a weighted channel-to-channel matrix, then uses a label-free symmetry score to pick the most informative head; on SEED-IV, classical classifiers reach 42.13 balanced accuracy with about 176 times fewer calculations than all-layer attention pooling.

The EEG foundation model REVE no longer needs full fine-tuning or all-layer pooled features to recognize emotion. Such models had not consistently beaten task-specific architectures on cross-subject tasks, and their large representations are costly to compute and store. This approach aggregates the first-layer attention weights of a single head over each trial into a weighted channel-to-channel matrix, then uses a label-free symmetry score to pick the most informative head; on SEED-IV, classical classifiers reach 42.13 balanced accuracy with about 176 times fewer calculations than all-layer attention pooling.

Addressing the gap in Speech Recognition Threshold (SRT) materials for Greek-speaking school-aged children (6 to 12 years), this study selected words on four criteria—syllabic structure (trisyllabics), age-appropriate vocabulary familiarity, phonemic differentiation, and homogeneity in audibility—recorded and processed them per ISO 8253-3:2022, had 20 children take part in the word homogeneity evaluation, determined for each word the presentation level needed for 50% correct recognition, and measured recognition rates across intensities to locate the steepest rise of the Performance-Intensity (PI) function between 20% and 80% recognition; it found that words homogeneous in recognition rate are not homogeneous in recognition threshold and vice versa, so only words within +1 Standard Deviati

Addressing the gap in Speech Recognition Threshold (SRT) materials for Greek-speaking school-aged children (6 to 12 years), this study selected words on four criteria—syllabic structure (trisyllabics), age-appropriate vocabulary familiarity, phonemic differentiation, and homogeneity in audibility—recorded and processed them per ISO 8253-3:2022, had 20 children take part in the word homogeneity evaluation, determined for each word the presentation level needed for 50% correct recognition, and measured recognition rates across intensities to locate the steepest rise of the Performance-Intensity (PI) function between 20% and 80% recognition; it found that words homogeneous in recognition rate are not homogeneous in recognition threshold and vice versa, so only words within +1 Standard Deviati