Training Strategies to Handle Missing Modalities for Audio-Visual Expression Recognition
Open accessThe DOI matches its Crossref record; 6 of the 6 references it cites were found in a registry or an index.
Abstract
Automatic audio-visual expression recognition can play an important role in communication services such as tele-health, VOIP calls and human-machine interaction. Accuracy of audio-visual expression recognition could benefit from the interplay between the two modalities. However, most audio-visual expression recognition systems, trained in ideal conditions, fail to generalize in real world scenarios where either the audio or visual modality could be missing due to a number of reasons such as limited bandwidth, interactors' orientation, caller initiated muting. This paper studies the performance of a state-of-the art transformer when one of the modalities is missing. We conduct ablation studies to evaluate the model in the absence of either modality. Further, we propose a strategy to randomly ablate visual inputs during training at the clip or frame level to mimic real world scenarios. Results conducted on in-the-wild data, indicate significant generalization in proposed models trained on missing cues, with gains up to 17% for frame level ablations, showing that these training strategies cope better with the loss of input modalities.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Lip reading sentences in the wild
Chung, J. S., Senior, A., Vinyals, O., & Zisserman, A. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.367 · Cited by 702 · cited 1×
“lipreading [4], action recognition [14], this study understands their dynamics for expression recognition.”
BackgroundExpression, Affect, Action Unit Recognition: Aff-Wild2, Multi-Task Learning and ArcFace
Kollias, D., & Zafeiriou, S. (2019) · arXiv · doi:10.48550/arxiv.1910.04855 · Cited by 210 · cited 2×
“Kollias et al. [13] provide a late fusion approach to concatenate audio and visual cues extracted through a deep convolutional neural network.”
BackgroundFear-type emotion recognition for future audio-based surveillance systems
Clavel, C., Vasilescu, I., Devillers, L., Richard, G., & Ehrette, T. (2008) · Speech Communication · doi:10.1016/j.specom.2008.03.012 · Cited by 180 · cited 1×
“Expression recognition systems have varied uses in fields such as health [6], security [5], customer interaction [9].”
BackgroundLearning supervised scoring ensemble for emotion recognition in the wild
Hu, P., Cai, D., Wang, S., Yao, A., & Chen, Y. (2017) · Proceedings of the 19th ACM International Conference on Multimodal Interaction · doi:10.1145/3136755.3143009 · Cited by 162
Analysis of physiological signals for recognition of boredom, pain, and surprise emotions
Jang, E.-H., Park, B.-J., Park, M.-S., Kim, S.-H., & Sohn, J.-H. (2015) · Journal of Physiological Anthropology · doi:10.1186/s40101-015-0063-5 · Cited by 159 · cited 1×
“In particular, expression recognition has interesting uses in communication services where systems can detect the onset of boredom [11], extract salient regions (hotspots) in an interaction [19] or indexing [21] Recently, the proliferation of audio and visual sensors has seen a rise in development…”
BackgroundProgressive Neural Networks for Transfer Learning in Emotion Recognition
Gideon, J., Khorram, S., Aldeneh, Z., Dimitriadis, D., & Provost, E. M. (2017) · Interspeech 2017 · doi:10.21437/interspeech.2017-1637 · Cited by 129
Cited by
Sun, L., Lian, Z., Liu, B., & Tao, J. (2024) · IEEE Transactions on Affective Computing · doi:10.1109/taffc.2023.3274829 · Cited by 258
Missing Modality Imagination Network for Emotion Recognition with Uncertain Missing Modalities
Zhao, J., Li, R., & Jin, Q. (2021) · Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) · doi:10.18653/v1/2021.acl-long.203 · Cited by 275
“Existing multimodal fusion models trained on full-modality samples usually fail when partial modalities are missing (Aguilar et al., 2019; Pham et al., 2019; Cai et al., 2018; Parthasarathy and Sundaram, 2020).”
BackgroundMethodInfluentialTag-assisted Multimodal Sentiment Analysis under Uncertain Missing Modalities
Zeng, J., Liu, T., & Zhou, J. (2022) · Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval · doi:10.1145/3477495.3532064 · Cited by 94
Liu, R., Zuo, H., Lian, Z., Schuller, B. W., & Li, H. (2024) · IEEE Transactions on Affective Computing · doi:10.1109/taffc.2024.3378570 · Cited by 96
Liu, J., Capurro, D., Nguyen, A., & Verspoor, K. (2023) · Journal of Biomedical Informatics · doi:10.1016/j.jbi.2023.104466 · Cited by 66
Metadata
- Authors
- Parthasarathy, S., & Sundaram, S.
- Venue
- Companion Publication of the 2020 International Conference on Multimodal Interaction
- Publisher
- ACM
- Published
- 2020-10-25
- Pages
- 400-404
- OpenAlex
- W3090765191
- Cited by
- 78
- References
- 6
- FWCI
- 8.44
- Type
- conference paper
- Language
- en
- Open access
- green
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.