COLD Fusion: Calibrated and Ordinal Latent Distribution Fusion for Uncertainty-Aware Multimodal Emotion Recognition
Open accessThe DOI matches its Crossref record; 36 of the 36 references it cites were found in a registry or an index.
Abstract
Automatically recognising apparent emotions from face and voice is hard, in part because of various sources of uncertainty, including in the input data and the labels used in a machine learning framework. This paper introduces an uncertainty-aware multimodal fusion approach that quantifies modality-wise aleatoric or data uncertainty towards emotion prediction. We propose a novel fusion framework, in which latent distributions over unimodal temporal context are learned by constraining their variance. These variance constraints, Calibration and Ordinal Ranking, are designed such that the variance estimated for a modality can represent how informative the temporal context of that modality is w.r.t. emotion recognition. When well-calibrated, modality-wise uncertainty scores indicate how much their corresponding predictions are likely to differ from the ground truth labels. Well-ranked uncertainty scores allow the ordinal ranking of different frames across different modalities. To jointly impose both these constraints, we propose a softmax distributional matching loss. Our evaluation on AVEC 2019 CES, CMU-MOSEI, and IEMOCAP datasets shows that the proposed multimodal fusion method not only improves the generalisation performance of emotion recognition models and their predictive uncertainty estimates, but also makes the models robust to novel noise patterns encountered at test time.
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
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946
A concordance correlation coefficient to evaluate reproducibility.
Lin, L. I.-K. (1989) · Biometrics · doi:10.2307/2532051 · Cited by 8,785 · cited 6×
“Regression models’ performance is measured using Lin’s Concordance Correlation Coefficient (CCC) [82] between the predicted emotions y o and their ground truth labels y where ρ y ∗ y o denotes the Pearson’s coefficient of correlation between y ∗ and y o , and ( μ y ∗ , μ y o ) and ( σ y ∗ , σ y o ) denote their mean and…”
MethodInfluentialSGDR: Stochastic Gradient Descent with Warm Restarts
Loshchilov, I., & Hutter, F. (2016) · arXiv · doi:10.48550/arxiv.1608.03983 · Cited by 5,309
Multimodal Machine Learning: A Survey and Taxonomy
Baltrusaitis, T., Ahuja, C., & Morency, L.-P. (2019) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2018.2798607 · Cited by 5,018
IEMOCAP: Interactive emotional dyadic motion capture database
Busso, C., Bulut, M., Lee, C.-C., Kazemzadeh, A., Mower, E., Kim, S., Chang, J. N., Lee, S., & Narayanan, S. S. (2008) · Language Resources and Evaluation · doi:10.1007/s10579-008-9076-6 · Cited by 3,869
SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition
Park, D. S., Chan, W., Zhang, Y., Chiu, C.-C., Zoph, B., Cubuk, E. D., & Le, Q. V. (2019) · Interspeech 2019 · doi:10.21437/interspeech.2019-2680 · Cited by 3,558
On calibration of modern neural networks
Guo, C., Pleiss, G., Sun, Y., & Weinberger, K. Q. (2017) · arXiv · doi:10.48550/arxiv.1706.04599 · Cited by 3,435
Deep Neural Networks are Easily Fooled: High Confidence Predictions for Unrecognizable Images
Nguyen, A., Yosinski, J., & Clune, J. (2014) · arXiv · doi:10.48550/arxiv.1412.1897 · Cited by 3,258
Deep neural networks are easily fooled: High confidence predictions for unrecognizable images
Nguyen, A., Yosinski, J., & Clune, J. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298640 · Cited by 3,258
A Survey of Affect Recognition Methods: Audio, Visual, and Spontaneous Expressions
Zeng, Z., Pantic, M., Roisman, G. I., & Huang, T. S. (2009) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2008.52 · Cited by 2,797
AffectNet: A Database for Facial Expression, Valence, and Arousal Computing in the Wild
Mollahosseini, A., Hasani, B., & Mahoor, M. H. (2019) · IEEE Transactions on Affective Computing · doi:10.1109/taffc.2017.2740923 · Cited by 2,338
Multimodal Transformer for Unaligned Multimodal Language Sequences
Tsai, Y.-H. H., Bai, S., Liang, P. P., Kolter, J. Z., Morency, L.-P., & Salakhutdinov, R. (2019) · Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/p19-1656 · Cited by 2,259 · cited 2×
“In Appendix A, available online, we compare the proposed COLD fusion and a multimodal Transformer baseline [99] on 8)).”
MethodTensor Fusion Network for Multimodal Sentiment Analysis
Zadeh, A., Chen, M., Poria, S., Cambria, E., & Morency, L.-P. (2017) · Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/d17-1115 · Cited by 2,102
Aleatoric and epistemic uncertainty in machine learning: an introduction to concepts and methods
Hüllermeier, E., & Waegeman, W. (2021) · Machine Learning · doi:10.1007/s10994-021-05946-3 · Cited by 1,834 · cited 1×
“In particular, we aim to estimate the aleatoric component of uncertainty [1] associated with different modalities for improved emotion recognition performance.”
BackgroundA review of affective computing: From unimodal analysis to multimodal fusion
Poria, S., Cambria, E., Bajpai, R., & Hussain, A. (2017) · Information Fusion · doi:10.1016/j.inffus.2017.02.003 · Cited by 1,644 · cited 1×
“We refer the reader to Poria et al. [30], Roust et al. [8], Jiang et al. [31] and Zhao et al. [32] for comprehensive surveys of affect recognition in multimodal settings and contemporary deep learning-specific advancements in it.”
BackgroundMultimodal Language Analysis in the Wild: CMU-MOSEI Dataset and Interpretable Dynamic Fusion Graph
Bagher Zadeh, A., Liang, P. P., Poria, S., Cambria, E., & Morency, L.-P. (2018) · Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · doi:10.18653/v1/p18-1208 · Cited by 1,433 · cited 3×
“We evaluate the proposed COLD fusion approach on: (a) dimensional emotion recognition from face and voice modalities in the AVEC 2019 CES [4] and IEMOCAP [20] datasets, and (b) categorical emotion recognition from face, voice and text modalities in the CMU-MOSEI [21] and IEMOCAP datasets.”
BackgroundMethodEfficient Low-rank Multimodal Fusion With Modality-Specific Factors
Liu, Z., Shen, Y., Lakshminarasimhan, V. B., Liang, P. P., Bagher Zadeh, A., & Morency, L.-P. (2018) · Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · doi:10.18653/v1/p18-1209 · Cited by 1,170
MISA: Modality-Invariant and -Specific Representations for Multimodal Sentiment Analysis
Hazarika, D., Zimmermann, R., & Poria, S. (2020) · Proceedings of the 28th ACM International Conference on Multimedia · doi:10.1145/3394171.3413678 · Cited by 1,073
Suppressing Uncertainties for Large-Scale Facial Expression Recognition
Wang, K., Peng, X., Yang, J., Lu, S., & Qiao, Y. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr42600.2020.00693 · Cited by 772
End-to-End Multimodal Emotion Recognition using Deep Neural Networks
Tzirakis, P., Trigeorgis, G., Nicolaou, M. A., Schuller, B., & Zafeiriou, S. (2017) · arXiv · doi:10.48550/arxiv.1704.08619 · Cited by 752 · cited 2×
“This method is also referred to as ‘feature fusion with RNNs’ or ‘mid-level’ fusion in some prior works [8], [72].”
BackgroundMethodSelf-citationPham, H., Liang, P. P., Manzini, T., Morency, L.-P., & Póczos, B. (2019) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v33i01.33016892 · Cited by 512
Nicolaou, M. A., Gunes, H., & Pantic, M. (2011) · IEEE Transactions on Affective Computing · doi:10.1109/t-affc.2011.9 · Cited by 450 · cited 1×
“We believe that uncertainty-aware multimodal fusion may have the potential to address this challenge, considering that the intensity of spontaneous emotions embedded in different modalities are likely to vary dynamically over time [6], [7].”
BackgroundSun, Z., Sarma, P., Sethares, W., & Liang, Y. (2020) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v34i05.6431 · Cited by 410
Ringeval, F., Schuller, B., Valstar, M., Cummins, N., Cowie, R., Tavabi, L., Schmitt, M., Alisamir, S., Amiriparian, S., Messner, E.-M., et al. (2019) · Proceedings of the 9th International on Audio/Visual Emotion Challenge and Workshop · doi:10.1145/3347320.3357688 · Cited by 409 · cited 6×
“Being an intrinsically temporal and multimodal phenomenon, emotion recognition from multimodal inputs is a long-standing challenge in Affective Computing [2], [3], [4].”
BackgroundMethodInfluentialSelf-citationLearning Affective Features With a Hybrid Deep Model for Audio–Visual Emotion Recognition
Zhang, S., Zhang, S., Huang, T., Gao, W., & Tian, Q. (2018) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2017.2719043 · Cited by 346 · cited 1×
“, [58]), not accounting for commonly encountered temporal misalignment between different modalities [59].”
BackgroundData Uncertainty Learning in Face Recognition
Chang, J., Lan, Z., Cheng, C., & Wei, Y. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr42600.2020.00575 · Cited by 320
Garnelo, M., Rosenbaum, D., Maddison, C. J., Ramalho, T., Saxton, D., Shanahan, M., Teh, Y. W., Rezende, D. J., & Eslami, S. M. A. (2018) · arXiv · doi:10.48550/arxiv.1807.01613 · Cited by 287 · cited 1×
“With the same motivation, in Affective Processes (APs) [40], [41], Neural Processes [42], [43] have been applied to the task of emotion recognition.”
BackgroundSewa db: A rich database for audio-visual emotion and sentiment research in the wild
Kossaifi, J., Walecki, R., Panagakis, Y., Shen, J., Schmitt, M., Ringeval, F., Han, J., Pandit, V., Toisoul, A., Schuller, B., et al. (2021) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2019.2944808 · Cited by 243
Self-citationDeep learning for human affect recognition: Insights and new developments
Rouast, P. V., Adam, M. T. P., & Chiong, R. (2021) · IEEE Transactions on Affective Computing · doi:10.1109/taffc.2018.2890471 · Cited by 243
Selective Classification for Deep Neural Networks
Geifman, Y., & El-Yaniv, R. (2017) · arXiv · doi:10.48550/arxiv.1705.08500 · Cited by 220
Addressing Failure Prediction by Learning Model Confidence
Corbière, C., Thome, N., Bar-Hen, A., Cord, M., & Pérez, P. (2019) · arXiv · doi:10.48550/arxiv.1910.04851 · Cited by 211 · cited 1×
“Other closely related works approached the problem of ordinal ranking of uncertainty estimates with different objectives suchasfailureprediction[65],out-of-distributiondetection[66], andselectiveclassification[67].”
BackgroundPrediction of asynchronous dimensional emotion ratings from audiovisual and physiological data
Ringeval, F., Eyben, F., Kroupi, E., Yuce, A., Thiran, J.-P., Ebrahimi, T., Lalanne, D., & Schuller, B. (2015) · Pattern Recognition Letters · doi:10.1016/j.patrec.2014.11.007 · Cited by 202 · cited 2×
“Decision Fusion combines the unimodal emotion predictions Y V and Y A (e.g., [71]).”
BackgroundResultSelf-citationMultimodal Multi-task Learning for Dimensional and Continuous Emotion Recognition
Chen, S., Jin, Q., Zhao, J., & Wang, S. (2017) · Proceedings of the 7th Annual Workshop on Audio/Visual Emotion Challenge · doi:10.1145/3133944.3133949 · Cited by 147 · cited 1×
“To differentiate the interlocutor’s information from that of the target speaker, we implemented the feature dimensionality-doubling technique proposed in [91].”
MethodMultimodal End-to-End Sparse Model for Emotion Recognition
Dai, W., Cahyawijaya, S., Liu, Z., & Fung, P. (2021) · Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies · doi:10.18653/v1/2021.naacl-main.417 · Cited by 91
Uncertainty-Aware Audiovisual Activity Recognition Using Deep Bayesian Variational Inference
Subedar, M., Krishnan, R., Meyer, P. L., Tickoo, O., & Huang, J. (2019) · 2019 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00640 · Cited by 85 · cited 1×
“[41] applied Bayesian DNNs for uncertainty-aware audiovisual fusion to improve human activity recognition performance.”
BackgroundLingenfelser, F., Wagner, J., Deng, J., Brueckner, R., Schuller, B., & Andre, E. (2018) · IEEE Transactions on Affective Computing · doi:10.1109/taffc.2016.2635124 · Cited by 23 · cited 2×
“This performance degradation may be due to not explicitly correcting the temporal misalignment effects [59], which are heuristically derived in general [3].”
BackgroundSelf-citation
Cited by
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
Self-citationMultimodal emotion recognition: A comprehensive review, trends, and challenges
Ramaswamy, M. P. A., & Palaniswamy, S. (2024) · WIREs Data Mining and Knowledge Discovery · doi:10.1002/widm.1563 · Cited by 70
Chumachenko, K., Iosifidis, A., & Gabbouj, M. (2024) · 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw63382.2024.00470 · Cited by 34
Metadata
- Authors
- Tellamekala, M. K., Amiriparian, S., Schuller, B. W., André, E., Giesbrecht, T., & Valstar, M.
- Venue
- IEEE Transactions on Pattern Analysis and Machine Intelligence
- Published
- 2024-02
- Volume
- 46
- Issue
- 2
- Pages
- 805-822
- ISSN
- 0162-8828 (ISSN-L)
- 2160-9292
- 1939-3539
- OpenAlex
- W4387757537
- Cited by
- 66
- References
- 36
- FWCI
- 15.33
- Type
- journal article
- 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.