Scaling Multimodal Pre-Training via Cross-Modality Gradient Harmonization

• URL: http://arxiv.org/abs/2211.02077v1
• Date: Thu, 3 Nov 2022 18:12:32 GMT
• Title: Scaling Multimodal Pre-Training via Cross-Modality Gradient Harmonization
• Authors: Junru Wu, Yi Liang, Feng Han, Hassan Akbari, Zhangyang Wang, Cong Yu
• Abstract summary: Self-supervised pre-training recently demonstrates success on large-scale multimodal data. Cross-modality alignment (CMA) is only a weak and noisy supervision. CMA might cause conflicts and biases among modalities.
• Score: 68.49738668084693
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Self-supervised pre-training recently demonstrates success on large-scale multimodal data, and state-of-the-art contrastive learning methods often enforce the feature consistency from cross-modality inputs, such as video/audio or video/text pairs. Despite its convenience to formulate and leverage in practice, such cross-modality alignment (CMA) is only a weak and noisy supervision, since two modalities can be semantically misaligned even they are temporally aligned. For example, even in the commonly adopted instructional videos, a speaker can sometimes refer to something that is not visually present in the current frame; and the semantic misalignment would only be more unpredictable for the raw videos from the internet. We conjecture that might cause conflicts and biases among modalities, and may hence prohibit CMA from scaling up to training with larger and more heterogeneous data. This paper first verifies our conjecture by observing that, even in the latest VATT pre-training using only instructional videos, there exist strong gradient conflicts between different CMA losses within the same video, audio, text triplet, indicating them as the noisy source of supervision. We then propose to harmonize such gradients, via two techniques: (i) cross-modality gradient realignment: modifying different CMA loss gradients for each sample triplet, so that their gradient directions are more aligned; and (ii) gradient-based curriculum learning: leveraging the gradient conflict information on an indicator of sample noisiness, to develop a curriculum learning strategy to prioritize training on less noisy sample triplets. Applying those techniques to pre-training VATT on the HowTo100M dataset, we consistently improve its performance on different downstream tasks. Moreover, we are able to scale VATT pre-training to more complicated non-narrative Youtube8M dataset to further improve the state-of-the-arts.

Related papers

• Two-Stage Triplet Loss Training with Curriculum Augmentation for Audio-Visual Retrieval [3.164991885881342]
  Cross- retrieval models learn robust embedding spaces. We introduce a novel approach rooted in curriculum learning to address this problem. We propose a two-stage training paradigm that guides the model's learning process from semi-hard to hard triplets.
  arXiv  Detail & Related papers  (2023-10-20T12:35:54Z)
• Cross-head mutual Mean-Teaching for semi-supervised medical image segmentation [6.738522094694818]
  Semi-supervised medical image segmentation (SSMIS) has witnessed substantial advancements by leveraging limited labeled data and abundant unlabeled data. Existing state-of-the-art (SOTA) methods encounter challenges in accurately predicting labels for the unlabeled data. We propose a novel Cross-head mutual mean-teaching Network (CMMT-Net) incorporated strong-weak data augmentation.
  arXiv  Detail & Related papers  (2023-10-08T09:13:04Z)
• Few-Shot Classification with Contrastive Learning [10.236150550121163]
  We propose a novel contrastive learning-based framework that seamlessly integrates contrastive learning into both stages. In the meta-training stage, we propose a cross-view episodic training mechanism to perform the nearest centroid classification on two different views of the same episode. These two strategies force the model to overcome the bias between views and promote the transferability of representations.
  arXiv  Detail & Related papers  (2022-09-17T02:39:09Z)
• PA-Seg: Learning from Point Annotations for 3D Medical Image Segmentation using Contextual Regularization and Cross Knowledge Distillation [14.412073730567137]
  We propose to annotate a segmentation target with only seven points in 3D medical images, and design a two-stage weakly supervised learning framework PA-Seg. In the first stage, we employ geodesic distance transform to expand the seed points to provide more supervision signal. In the second stage, we use predictions obtained by the model pre-trained in the first stage as pseudo labels.
  arXiv  Detail & Related papers  (2022-08-11T07:00:33Z)
• VLMixer: Unpaired Vision-Language Pre-training via Cross-Modal CutMix [59.25846149124199]
  This paper proposes a data augmentation method, namely cross-modal CutMix. CMC transforms natural sentences from the textual view into a multi-modal view. By attaching cross-modal noise on uni-modal data, it guides models to learn token-level interactions across modalities for better denoising.
  arXiv  Detail & Related papers  (2022-06-17T17:56:47Z)
• On Modality Bias Recognition and Reduction [70.69194431713825]
  We study the modality bias problem in the context of multi-modal classification. We propose a plug-and-play loss function method, whereby the feature space for each label is adaptively learned. Our method yields remarkable performance improvements compared with the baselines.
  arXiv  Detail & Related papers  (2022-02-25T13:47:09Z)
• Dense Contrastive Visual-Linguistic Pretraining [53.61233531733243]
  Several multimodal representation learning approaches have been proposed that jointly represent image and text. These approaches achieve superior performance by capturing high-level semantic information from large-scale multimodal pretraining. We propose unbiased Dense Contrastive Visual-Linguistic Pretraining to replace the region regression and classification with cross-modality region contrastive learning.
  arXiv  Detail & Related papers  (2021-09-24T07:20:13Z)
• Contrastive Prototype Learning with Augmented Embeddings for Few-Shot Learning [58.2091760793799]
  We propose a novel contrastive prototype learning with augmented embeddings (CPLAE) model. With a class prototype as an anchor, CPL aims to pull the query samples of the same class closer and those of different classes further away. Extensive experiments on several benchmarks demonstrate that our proposed CPLAE achieves new state-of-the-art.
  arXiv  Detail & Related papers  (2021-01-23T13:22:44Z)
• Recurrent Multi-view Alignment Network for Unsupervised Surface Registration [79.72086524370819]
  Learning non-rigid registration in an end-to-end manner is challenging due to the inherent high degrees of freedom and the lack of labeled training data. We propose to represent the non-rigid transformation with a point-wise combination of several rigid transformations. We also introduce a differentiable loss function that measures the 3D shape similarity on the projected multi-view 2D depth images.
  arXiv  Detail & Related papers  (2020-11-24T14:22:42Z)

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.