We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

eess.AS

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Electrical Engineering and Systems Science > Audio and Speech Processing

Title: Detecting expressions with multimodal transformers

Abstract: Developing machine learning algorithms to understand person-to-person engagement can result in natural user experiences for communal devices such as Amazon Alexa. Among other cues such as voice activity and gaze, a person's audio-visual expression that includes tone of the voice and facial expression serves as an implicit signal of engagement between parties in a dialog. This study investigates deep-learning algorithms for audio-visual detection of user's expression. We first implement an audio-visual baseline model with recurrent layers that shows competitive results compared to current state of the art. Next, we propose the transformer architecture with encoder layers that better integrate audio-visual features for expressions tracking. Performance on the Aff-Wild2 database shows that the proposed methods perform better than baseline architecture with recurrent layers with absolute gains approximately 2% for arousal and valence descriptors. Further, multimodal architectures show significant improvements over models trained on single modalities with gains of up to 3.6%. Ablation studies show the significance of the visual modality for the expression detection on the Aff-Wild2 database.
Comments: IEEE Spoken Language Technology Workshop 2021
Subjects: Audio and Speech Processing (eess.AS); Sound (cs.SD); Image and Video Processing (eess.IV)
Cite as: arXiv:2012.00063 [eess.AS]
  (or arXiv:2012.00063v1 [eess.AS] for this version)

Submission history

From: Srinivas Parthasarathy [view email]
[v1] Mon, 30 Nov 2020 19:31:03 GMT (2806kb,D)

Link back to: arXiv, form interface, contact.