Mind the Gap: Understanding the Modality Gap in Multi-modal Contrastive Representation Learning

• URL: http://arxiv.org/abs/2203.02053v1
• Date: Thu, 3 Mar 2022 22:53:54 GMT
• Title: Mind the Gap: Understanding the Modality Gap in Multi-modal Contrastive Representation Learning
• Authors: Weixin Liang, Yuhui Zhang, Yongchan Kwon, Serena Yeung, James Zou
• Abstract summary: We show that different data modalities are embedded at arm's length in their shared representation in multi-modal models such as CLIP. In contrastive learning keeps the different modalities separate by a certain distance, which is influenced by the temperature parameter in the loss function. Our experiments further demonstrate that varying the modality gap distance has a significant impact in improving the model's downstream zero-shot classification performance and fairness.
• Score: 35.25854322376364
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: We present modality gap, an intriguing geometric phenomenon of the representation space of multi-modal models. Specifically, we show that different data modalities (e.g. images and text) are embedded at arm's length in their shared representation in multi-modal models such as CLIP. Our systematic analysis demonstrates that this gap is caused by a combination of model initialization and contrastive learning optimization. In model initialization, we show empirically and theoretically that the representation of a common deep neural network is restricted to a narrow cone. As a consequence, in a multi-modal model with two encoders, the representations of the two modalities are clearly apart when the model is initialized. During optimization, contrastive learning keeps the different modalities separate by a certain distance, which is influenced by the temperature parameter in the loss function. Our experiments further demonstrate that varying the modality gap distance has a significant impact in improving the model's downstream zero-shot classification performance and fairness. Our code and data are available at https://modalitygap.readthedocs.io/

Related papers

• Human-Object Interaction Detection Collaborated with Large Relation-driven Diffusion Models [65.82564074712836]
  We introduce DIFfusionHOI, a new HOI detector shedding light on text-to-image diffusion models. We first devise an inversion-based strategy to learn the expression of relation patterns between humans and objects in embedding space. These learned relation embeddings then serve as textual prompts, to steer diffusion models generate images that depict specific interactions.
  arXiv  Detail & Related papers  (2024-10-26T12:00:33Z)
• Semantic-Guided Multimodal Sentiment Decoding with Adversarial Temporal-Invariant Learning [22.54577327204281]
  Multimodal sentiment analysis aims to learn representations from different modalities to identify human emotions. Existing works often neglect the frame-level redundancy inherent in continuous time series, resulting in incomplete modality representations with noise. We propose temporal-invariant learning for the first time, which constrains the distributional variations over time steps to effectively capture long-term temporal dynamics.
  arXiv  Detail & Related papers  (2024-08-30T03:28:40Z)
• Debiasing Multimodal Models via Causal Information Minimization [65.23982806840182]
  We study bias arising from confounders in a causal graph for multimodal data. Robust predictive features contain diverse information that helps a model generalize to out-of-distribution data. We use these features as confounder representations and use them via methods motivated by causal theory to remove bias from models.
  arXiv  Detail & Related papers  (2023-11-28T16:46:14Z)
• Improving Discriminative Multi-Modal Learning with Large-Scale Pre-Trained Models [51.5543321122664]
  This paper investigates how to better leverage large-scale pre-trained uni-modal models to enhance discriminative multi-modal learning. We introduce Multi-Modal Low-Rank Adaptation learning (MMLoRA)
  arXiv  Detail & Related papers  (2023-10-08T15:01:54Z)
• UniDiff: Advancing Vision-Language Models with Generative and Discriminative Learning [86.91893533388628]
  This paper presents UniDiff, a unified multi-modal model that integrates image-text contrastive learning (ITC), text-conditioned image synthesis learning (IS), and reciprocal semantic consistency modeling (RSC) UniDiff demonstrates versatility in both multi-modal understanding and generative tasks.
  arXiv  Detail & Related papers  (2023-06-01T15:39:38Z)
• Characterizing and overcoming the greedy nature of learning in multi-modal deep neural networks [62.48782506095565]
  We show that due to the greedy nature of learning in deep neural networks, models tend to rely on just one modality while under-fitting the other modalities. We propose an algorithm to balance the conditional learning speeds between modalities during training and demonstrate that it indeed addresses the issue of greedy learning.
  arXiv  Detail & Related papers  (2022-02-10T20:11:21Z)
• Discriminative Multimodal Learning via Conditional Priors in Generative Models [21.166519800652047]
  This research studies the realistic scenario in which all modalities and class labels are available for model training. We show, in this scenario, that the variational lower bound limits mutual information between joint representations and missing modalities.
  arXiv  Detail & Related papers  (2021-10-09T17:22:24Z)
• MOCCA: Multi-Layer One-Class ClassificAtion for Anomaly Detection [16.914663209964697]
  We propose our deep learning approach to the anomaly detection problem named Multi-LayerOne-Class Classification (MOCCA) We explicitly leverage the piece-wise nature of deep neural networks by exploiting information extracted at different depths to detect abnormal data instances. We show that our method reaches superior performances compared to the state-of-the-art approaches available in the literature.
  arXiv  Detail & Related papers  (2020-12-09T08:32:56Z)
• Does my multimodal model learn cross-modal interactions? It's harder to tell than you might think! [26.215781778606168]
  Cross-modal modeling seems crucial in multimodal tasks, such as visual question answering. We propose a new diagnostic tool, empirical multimodally-additive function projection (EMAP), for isolating whether or not cross-modal interactions improve performance for a given model on a given task. For seven image+text classification tasks (on each of which we set new state-of-the-art benchmarks), we find that, in many cases, removing cross-modal interactions results in little to no performance degradation.
  arXiv  Detail & Related papers  (2020-10-13T17:45:28Z)
• Learning more expressive joint distributions in multimodal variational methods [0.17188280334580194]
  We introduce a method that improves the representational capacity of multimodal variational methods using normalizing flows. We demonstrate that the model improves on state-of-the-art multimodal methods based on variational inference on various computer vision tasks. We also show that learning more powerful approximate joint distributions improves the quality of the generated samples.
  arXiv  Detail & Related papers  (2020-09-08T11:45:27Z)
• Adaptive Context-Aware Multi-Modal Network for Depth Completion [107.15344488719322]
  We propose to adopt the graph propagation to capture the observed spatial contexts. We then apply the attention mechanism on the propagation, which encourages the network to model the contextual information adaptively. Finally, we introduce the symmetric gated fusion strategy to exploit the extracted multi-modal features effectively. Our model, named Adaptive Context-Aware Multi-Modal Network (ACMNet), achieves the state-of-the-art performance on two benchmarks.
  arXiv  Detail & Related papers  (2020-08-25T06:00:06Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.