Utterance Level Feature Aggregation with Deep Metric Learning for Speech Emotion Recognition

Bogdan Mocanu, Ruxandra Țapu, Titus B. Zaharia

Sensors 2021 article cited by 23

Source Open-access PDF More in Software & Coding

importance 4/10 ai assisted Emotion and Mood Recognition speech technology stage: research

Problem

Automatically recognizing emotions in human speech at the utterance level, needed for applications from mental disease diagnosis to human-robot interaction.

Why it matters

Emotion-aware speech systems matter for healthcare, surveillance, e-learning and human-machine interaction.

What the AI did

A SE-ResNet CNN fed with spectrograms is extended with a trainable GhostVLAD clustering layer and an emotionally constrained triplet loss to build robust utterance-level feature representations.

Results

not stated in source

Impact

Proposes a method to overcome weak utterance-level feature representation in state-of-the-art speech emotion recognition.

Record

Venue	Sensors
Published	2021
Work type	article
Authors	Bogdan Mocanu, Ruxandra Țapu, Titus B. Zaharia
Citations	23 (OpenAlex, at import time)
References	50
Model or method	SE-ResNet with GhostVLAD layer and triplet loss
Organization	not stated in source
Primary source	https://doi.org/10.3390/s21124233
Open access	gold
Keywords	speech emotion recognition, deep metric learning, SE-ResNet, GhostVLAD, triplet loss, spectrograms

Importance rating

4 / 10

real but narrower advance

Speech emotion recognition supports applications from mental disease diagnosis to human-robot interaction; a methodological advance on public datasets with no stated real-world deployment or scale.

Basis: inferred. Judged from the source text, which gives no explicit scale. Stated scale: not stated in source. Method and limits.

More in Software & Coding

#	Paper	Year	Citations	Importance
1	ImageNet classification with deep convolutional neural networks Alex Krizhevsky, Ilya Sutskever, Geoffrey E. Hinton	2017	106,345	8
2	Deep Residual Learning for Image Recognition Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun	2016	221,134	8
3	Faster sorting algorithms discovered using deep reinforcement learning Daniel J. Mankowitz, Andrea Michi, Anton Zhernov, Marco Gelmi, Marco S...	2023	175	8
4	A deep learning‐based approach for software vulnerability detection using code metrics Fazli Subhan, Xiaoxue Wu, Lili Bo, Xiaobing Sun, Muhammad Rahman	2022	42	7
5	Deep Learning-Based Attack Detection and Classification in Android Devices Alfonso Gómez, Antonio Muñoz	2023	41	7