Grupo de Tecnología del Habla y Aprendizaje Automático

Speech Technology and Machine Learning Group

A Proposal for Multimodal Emotion Recognition Using Aural Transformers and Action Units on RAVDESS Dataset

We keep boosting and improving the performance of our Multimodal Emotion Recognition solutions! Check our most recent advances in our new paper in Applied Sciences!

THAU wins at 4th edition of the Predicting Media Memorability Challenge hosted by the MediaEval Workshop (December 13-15th, 2021)

Out of the five main results required, our team achieved the best score in two of them while matching the state of the art in another two. Our proposal relied on extracting semantically-rich, non-adapted visual and textual features using Transformers, hence exploiting the correlation between semantic units and their intrinsic memorability.

Official results presentation video available at YouTube

Welcome to THAU

Welcome to the Speech Technology and Machine Learning group website

Guided Spatial Transformers for Facial Expression Recognition

Check our new paper published in Applied Sciences

Fine-Tuning BERT Models for Intent Recognition Using a Frequency Cut-Off Strategy for Domain-Specific Vocabulary Extension

In this work we analyze the effect of inserting unseen domain words to extend the vocabulary of a BERT model. Particularly, a very simple word frequency cut-off strategy is experimentally shown to be a suitable method for driving the vocabulary learning decisions over unseen words. Our results show that the proposed method helps to effectively extend the original vocabulary of the pretrained models. Check our new paper published in Applied Sciences

Research

Automatic Speech Recognition

Speaker Identification / Diarization

Language Recognition

Multimodal Emotion Recognition

Machine Translation

Keyword Spotting

Natural Language Understanding

Intent Detection

Spoken Dialog Systems

Conversational Agents (chatbots)

Text-To-Speech Synthesis

Topic identification

Text Generation

Trustworthiness Assessment

Wearable Sensing

Multimedia Information Retrieval

Image and Video Processing

Scene Understanding

Multimedia Content Perception

Human Motion Modeling

Recent

Estudio sobre sistemas de redacción automática de contenido

Desarrollo de un sistema de reconocimiento de idioma y de diarización a partir de la señal de habla

Sistemas conversacionales inteligentes dirigidos por objetivos, multimodales y basados en microservicios