An official website of the United States government
Learning multimodal representations is a fundamentally complex research problem due to the presence of multiple heterogeneous sources of information. Although the presence of multiple modalities provides additional valuable information, there are two key challenges to address when learning from multimodal data: 1) models must learn the complex intra-modal and cross-modal interactions for prediction and 2) models must be robust to unexpected missing or noisy modalities during testing. In this paper, we propose to optimize for a joint generative-discriminative objective across multimodal data and labels. We introduce a model that factorizes representations into two sets of independent factors: multimodal discriminative and modality-specific generative factors. Multimodal discriminative factors are shared across all modalities and contain joint multimodal features required for discriminative tasks such as sentiment prediction. Modality-specific generative factors are unique for each modality and contain the information required for generating data. Experimental results show that our model is able to learn meaningful multimodal representations that achieve state-of-the-art or competitive performance on six multimodal datasets. Our model demonstrates flexible generative capabilities by conditioning on independent factors and can reconstruct missing modalities without significantly impacting performance. Lastly, we interpret our factorized representations to understand the interactions that influence multimodal learning.
more »
« less
Neuro-Inspired Information-Theoretic Hierarchical Perception for Multimodal Learning
Integrating and processing information from various sources or modalities are critical for obtaining a comprehensive and accurate perception of the real world in autonomous systems and cyber-physical systems. Drawing inspiration from neuroscience, we develop the Information-Theoretic Hierarchical Perception (ITHP) model, which utilizes the concept of information bottleneck. Different from most traditional fusion models that incorporate all modalities identically in neural networks, our model designates a prime modality and regards the remaining modalities as detectors in the information pathway, serving to distill the flow of information. Our proposed perception model focuses on constructing an effective and compact information flow by achieving a balance between the minimization of mutual information between the latent state and the input modal state, and the maximization of mutual information between the latent states and the remaining modal states. This approach leads to compact latent state representations that retain relevant information while minimizing redundancy, thereby substantially enhancing the performance of multimodal representation learning. Experimental evaluations on the MUStARD, CMU-MOSI, and CMU-MOSEI datasets demonstrate that our model consistently distills crucial information in multimodal learning scenarios, outperforming state-of-the-art benchmarks. Remarkably, on the CMU-MOSI dataset, ITHP surpasses human-level performance in the multimodal sentiment binary classification task across all evaluation metrics (i.e., Binary Accuracy, F1 Score, Mean Absolute Error, and Pearson Correlation).
more »
« less
- PAR ID:
- 10528298
- Editor(s):
- Kim, Been
- Publisher / Repository:
- Twelfth International Conference on Learning Representations (ICLR)
- Date Published:
- Edition / Version:
- 1
- Volume:
- 2024
- Issue:
- 1
- ISBN:
- 9781713872740
- Page Range / eLocation ID:
- 1-29
- Subject(s) / Keyword(s):
- multimodal learning, information bottleneck, sentiment analysis
- Format(s):
- Medium: X Size: 1.5 Other: pdf
- Size(s):
- 1.5
- Location:
- Vienna, Austria
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Multimodal sentiment analysis is a core research area that studies speaker sentiment expressed from the language, visual, and acoustic modalities. The central challenge in multimodal learning involves inferring joint representations that can process and relate information from these modalities. However, existing work learns joint representations by requiring all modalities as input and as a result, the learned representations may be sensitive to noisy or missing modalities at test time. With the recent success of sequence to sequence (Seq2Seq) models in machine translation, there is an opportunity to explore new ways of learning joint representations that may not require all input modalities at test time. In this paper, we propose a method to learn robust joint representations by translating between modalities. Our method is based on the key insight that translation from a source to a target modality provides a method of learning joint representations using only the source modality as input. We augment modality translations with a cycle consistency loss to ensure that our joint representations retain maximal information from all modalities. Once our translation model is trained with paired multimodal data, we only need data from the source modality at test time for final sentiment prediction. This ensures that our model remains robust from perturbations or missing information in the other modalities. We train our model with a coupled translationprediction objective and it achieves new state-of-the-art results on multimodal sentiment analysis datasets: CMU-MOSI, ICTMMMO, and YouTube. Additional experiments show that our model learns increasingly discriminative joint representations with more input modalities while maintaining robustness to missing or perturbed modalities.more » « less
-
Human state recognition is a critical topic with pervasive and important applications in human–machine systems. Multimodal fusion, which entails integrating metrics from various data sources, has proven to be a potent method for boosting recognition performance. Although recent multimodal-based models have shown promising results, they often fall short in fully leveraging sophisticated fusion strategies essential for modeling adequate cross-modal dependencies in the fusion representation. Instead, they rely on costly and inconsistent feature crafting and alignment. To address this limitation, we propose an end-to-end multimodal transformer framework for multimodal human state recognition called Husformer. Specifically, we propose using cross-modal transformers, which inspire one modality to reinforce itself through directly attending to latent relevance revealed in other modalities, to fuse different modalities while ensuring sufficient awareness of the cross-modal interactions introduced. Subsequently, we utilize a self-attention transformer to further prioritize contextual information in the fusion representation. Extensive experiments on two human emotion corpora (DEAP and WESAD) and two cognitive load datasets [multimodal dataset for objective cognitive workload assessment on simultaneous tasks (MOCAS) and CogLoad] demonstrate that in the recognition of the human state, our Husformer outperforms both state-of-the-art multimodal baselines and the use of a single modality by a large margin, especially when dealing with raw multimodal features. We also conducted an ablation study to show the benefits of each component in Husformer. Experimental details and source code are available at https://github.com/SMARTlab-Purdue/Husformer.more » « less
-
Abstract Multimodal single-cell sequencing technologies provide unprecedented information on cellular heterogeneity from multiple layers of genomic readouts. However, joint analysis of two modalities without properly handling the noise often leads to overfitting of one modality by the other and worse clustering results than vanilla single-modality analysis. How to efficiently utilize the extra information from single cell multi-omics to delineate cell states and identify meaningful signal remains as a significant computational challenge. In this work, we propose a deep learning framework, named SAILERX, for efficient, robust, and flexible analysis of multi-modal single-cell data. SAILERX consists of a variational autoencoder with invariant representation learning to correct technical noises from sequencing process, and a multimodal data alignment mechanism to integrate information from different modalities. Instead of performing hard alignment by projecting both modalities to a shared latent space, SAILERX encourages the local structures of two modalities measured by pairwise similarities to be similar. This strategy is more robust against overfitting of noises, which facilitates various downstream analysis such as clustering, imputation, and marker gene detection. Furthermore, the invariant representation learning part enables SAILERX to perform integrative analysis on both multi- and single-modal datasets, making it an applicable and scalable tool for more general scenarios.more » « less
-
Recent Transformer-based contextual word representations, including BERT and XLNet, have shown state-of-the-art performance in multiple disciplines within NLP. Fine-tuning the trained contextual models on task-specific datasets has been the key to achieving superior performance downstream. While fine-tuning these pre-trained models is straightforward for lexical applications (applications with only language modality), it is not trivial for multimodal language (a growing area in NLP focused on modeling face-to-face communication). More specifically, this is due to the fact that pre-trained models don’t have the necessary components to accept two extra modalities of vision and acoustic. In this paper, we proposed an attachment to BERT and XLNet called Multimodal Adaptation Gate (MAG). MAG allows BERT and XLNet to accept multimodal nonverbal data during fine-tuning. It does so by generating a shift to internal representation of BERT and XLNet; a shift that is conditioned on the visual and acoustic modalities. In our experiments, we study the commonly used CMU-MOSI and CMU-MOSEI datasets for multimodal sentiment analysis. Fine-tuning MAG-BERT and MAG-XLNet significantly boosts the sentiment analysis performance over previous baselines as well as language-only fine-tuning of BERT and XLNet. On the CMU-MOSI dataset, MAG-XLNet achieves human-level multimodal sentiment analysis performance for the first time in the NLP community.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
