skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


Title: Jointly Learning From Unimodal and Multimodal-Rated Labels in Audio-Visual Emotion Recognition
Audio-visual emotion recognition (AVER) has been an important research area in human-computer interaction (HCI). Traditionally, audio-visual emotional datasets and corresponding models derive their ground truths from annotations obtained by raters after watching the audio-visual stimuli. This conventional method, however, neglects the nuanced human perception of emotional states, which varies when annotations are made under different emotional stimuli conditions—whether through unimodal or multimodal stimuli. This study investigates the potential for enhanced AVER system performance by integrating diverse levels of annotation stimuli, reflective of varying perceptual evaluations. We propose a two-stage training method to train models with the labels elicited by audio-only, face-only, and audio-visual stimuli. Our approach utilizes different levels of annotation stimuli according to which modality is present within different layers of the model, effectively modeling annotation at the unimodal and multi-modal levels to capture the full scope of emotion perception across unimodal and multimodal contexts. We conduct the experiments and evaluate the models on the CREMA-D emotion database. The proposed methods achieved the best performances in macro-/weighted-F1 scores. Additionally, we measure the model calibration, performance bias, and fairness metrics considering the age, gender, and race of the AVER systems.  more » « less
Award ID(s):
2016719
PAR ID:
10655465
Author(s) / Creator(s):
 ;  ;  ;  ;  
Publisher / Repository:
IEEE
Date Published:
Journal Name:
IEEE Open Journal of Signal Processing
Volume:
6
ISSN:
2644-1322
Page Range / eLocation ID:
165 to 174
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; ; ; ; (, IEEE)
    na (Ed.)
    In the field of affective computing, emotional annotations are highly important for both the recognition and synthesis of human emotions. Researchers must ensure that these emotional labels are adequate for modeling general human perception. An unavoidable part of obtaining such labels is that human annotators are exposed to known and unknown stimuli before and during the annotation process that can affect their perception. Emotional stimuli cause an affective priming effect, which is a pre-conscious phenomenon in which previous emotional stimuli affect the emotional perception of a current target stimulus. In this paper, we use sequences of emotional annotations during a perceptual evaluation to study the effect of affective priming on emotional ratings of speech. We observe that previous emotional sentences with extreme emotional content push annotations of current samples to the same extreme. We create a sentence-level bias metric to study the effect of affective priming on speech emotion recognition (SER) modeling. The metric is used to identify subsets in the database with more affective priming bias intentionally creating biased datasets. We train and test SER models using the full and biased datasets. Our results show that although the biased datasets have low inter-evaluator agreements, SER models for arousal and dominance trained with those datasets perform the best. For valence, the models trained with the less-biased datasets perform the best. 
    more » « less
  2. ; ; (, IEEE)
    Emotion recognition is inherently a multimodal problem. Humans use both audible and visual cues to determine a person’s emotions. There has been extensive improvement in the methods we use to fuse audio and visual representations between two unimodal deep-learning models. However, there is a lack of accommodation for modalities that have a disparity in the amount of computational resources needed to provide the same amount of temporal information. As the sequence length increases, current methods often make simplifications such as discarding frames or cropping the sequence. This paper introduces a chunking methodology designed for cross-attention-based multimodal transformer architectures. The approach involves segmenting the visual input—the more computationally demanding modality—into chunks. Cross-attention is then performed between the encoded audio and visual features instead of the original sequence lengths of the unimodal backbones. Our method achieves significant improvements over conventional cross-attention techniques in the audio-visual domain for a six-class emotional recognition problem, demonstrating better F1 score, precision, and recall on the CREMA-D database while reducing computational overhead. 
    more » « less
  3. ; ; (, Interspeech)
    Training emotion recognition models has relied heavily on human annotated data, which present diversity, quality, and cost challenges. In this paper, we explore the potential of Large Language Models (LLMs), specifically GPT-4, in automating or assisting emotion annotation. We compare GPT-4 with supervised models and/or humans in three aspects: agreement with human annotations, alignment with human perception, and impact on model training. We find that common metrics that use aggregated human annotations as ground truth can underestimate GPT-4's performance, and our human evaluation experiment reveals a consistent preference for GPT-4 annotations over humans across multiple datasets and evaluators. Further, we investigate the impact of using GPT-4 as an annotation filtering process to improve model training. Together, our findings highlight the great potential of LLMs in emotion annotation tasks and underscore the need for refined evaluation methodologies. 
    more » « less
  4. ; ; ; ; ; ; (, IEEE Transactions on Affective Computing)
    Emotional annotation of data is important in affective computing for the analysis, recognition, and synthesis of emotions. As raters perceive emotion, they make relative comparisons with what they previously experienced, creating “anchors” that influence the annotations. This unconscious influence of the emotional content of previous stimuli in the perception of emotions is referred to as the affective priming effect. This phenomenon is also expected in annotations conducted with out-of-order segments, a common approach for annotating emotional databases. Can the affective priming effect introduce bias in the labels? If yes, how does this bias affect emotion recognition systems trained with these labels? This study presents a detailed analysis of the affective priming effect and its influence on speech emotion recognition (SER). The analysis shows that the affective priming effect affects emotional attributes and categorical emotion annotations. We observe that if annotators assign an extreme score to previous sentences for an emotional attribute (valence, arousal, or dominance), they will tend to annotate the next sentence closer to that extreme. We conduct SER experiments using the most biased sentences. We observe that models trained on the biased sentences perform the best and have the lowest prediction uncertainty. 
    more » « less
  5. ; ; (, Sensors)
    Decades of scientific research have been conducted on developing and evaluating methods for automated emotion recognition. With exponentially growing technology, there is a wide range of emerging applications that require emotional state recognition of the user. This paper investigates a robust approach for multimodal emotion recognition during a conversation. Three separate models for audio, video and text modalities are structured and fine-tuned on the MELD. In this paper, a transformer-based crossmodality fusion with the EmbraceNet architecture is employed to estimate the emotion. The proposed multimodal network architecture can achieve up to 65% accuracy, which significantly surpasses any of the unimodal models. We provide multiple evaluation techniques applied to our work to show that our model is robust and can even outperform the state-of-the-art models on the MELD. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email