Speech Emotion Recognition On Crema D

Metrics

Accuracy

Results

Performance results of various models on this benchmark

		Paper Title
Vertically long patch ViT	94.07	Accuracy enhancement method for speech emotion recognition from spectrogram using temporal frequency correlation and positional information learning through knowledge transfer
ConformerXL-P	88.2	BigSSL: Exploring the Frontier of Large-Scale Semi-Supervised Learning for Automatic Speech Recognition
CoordViT	82.96	CoordViT: A Novel Method of Improve Vision Transformer-Based Speech Emotion Recognition using Coordinate Information Concatenate
SepTr + LeRaC	70.95	Learning Rate Curriculum
SepTr	70.47	SepTr: Separable Transformer for Audio Spectrogram Processing
ResNet-18 + SPEL	68.12	Self-paced ensemble learning for speech and audio classification
ViT	67.81	AST: Audio Spectrogram Transformer
ResNet-18 + PyNADA	65.15	Non-linear Neurons with Human-like Apical Dendrite Activations
GRU	55.01	Visually Guided Self Supervised Learning of Speech Representations

0 of 9 row(s) selected.