MADTP: Multimodal Alignment-Guided Dynamic Token Pruning for Accelerating Vision-Language Transformer

• URL: http://arxiv.org/abs/2403.02991v1
• Date: Tue, 5 Mar 2024 14:13:50 GMT
• Title: MADTP: Multimodal Alignment-Guided Dynamic Token Pruning for Accelerating Vision-Language Transformer
• Authors: Jianjian Cao and Peng Ye and Shengze Li and Chong Yu and Yansong Tang and Jiwen Lu and Tao Chen
• Abstract summary: Vision-Language Transformers (VLTs) have shown great success recently, but are accompanied by heavy computation costs. We propose a novel framework named Multimodal Alignment-Guided Dynamic Token Pruning (MADTP) for accelerating various VLTs.
• Score: 66.71930982549028
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Vision-Language Transformers (VLTs) have shown great success recently, but are meanwhile accompanied by heavy computation costs, where a major reason can be attributed to the large number of visual and language tokens. Existing token pruning research for compressing VLTs mainly follows a single-modality-based scheme yet ignores the critical role of aligning different modalities for guiding the token pruning process, causing the important tokens for one modality to be falsely pruned in another modality branch. Meanwhile, existing VLT pruning works also lack the flexibility to dynamically compress each layer based on different input samples. To this end, we propose a novel framework named Multimodal Alignment-Guided Dynamic Token Pruning (MADTP) for accelerating various VLTs. Specifically, we first introduce a well-designed Multi-modality Alignment Guidance (MAG) module that can align features of the same semantic concept from different modalities, to ensure the pruned tokens are less important for all modalities. We further design a novel Dynamic Token Pruning (DTP) module, which can adaptively adjust the token compression ratio in each layer based on different input instances. Extensive experiments on various benchmarks demonstrate that MADTP significantly reduces the computational complexity of kinds of multimodal models while preserving competitive performance. Notably, when applied to the BLIP model in the NLVR2 dataset, MADTP can reduce the GFLOPs by 80% with less than 4% performance degradation.

Related papers

• MULTIFLOW: Shifting Towards Task-Agnostic Vision-Language Pruning [28.254318215697527]
  Vision-Language models (VLMs) come with high computational costs due to their large number of parameters. Existing techniques for VLMs are task-specific, and thus require pruning the network from scratch for each new task of interest. We explore a new direction: Task-Agnostic Vision-Language Pruning (TA-language) We propose Multimodal FlowPruning (MULTIFLOW), a first, gradient-free, pruning framework for TA-language.
  arXiv  Detail & Related papers  (2024-04-08T15:51:21Z)
• Bi-directional Adapter for Multi-modal Tracking [67.01179868400229]
  We propose a novel multi-modal visual prompt tracking model based on a universal bi-directional adapter. We develop a simple but effective light feature adapter to transfer modality-specific information from one modality to another. Our model achieves superior tracking performance in comparison with both the full fine-tuning methods and the prompt learning-based methods.
  arXiv  Detail & Related papers  (2023-12-17T05:27:31Z)
• Lightweight In-Context Tuning for Multimodal Unified Models [57.10831399642176]
  MultiModal In-conteXt Tuning (M$2$IXT) is a lightweight module to enhance the ICL capabilities of multimodal unified models. When tuned on as little as 50K multimodal data, M$2$IXT can boost the few-shot ICL performance significantly.
  arXiv  Detail & Related papers  (2023-10-08T10:47:24Z)
• Learned Thresholds Token Merging and Pruning for Vision Transformers [5.141687309207561]
  This paper introduces Learned Thresholds token Merging and Pruning (LTMP), a novel approach that leverages the strengths of both token merging and token pruning. We demonstrate our approach with extensive experiments on vision transformers on the ImageNet classification task.
  arXiv  Detail & Related papers  (2023-07-20T11:30:12Z)
• FM-ViT: Flexible Modal Vision Transformers for Face Anti-Spoofing [88.6654909354382]
  We present a pure transformer-based framework, dubbed the Flexible Modal Vision Transformer (FM-ViT) for face anti-spoofing. FM-ViT can flexibly target any single-modal (i.e., RGB) attack scenarios with the help of available multi-modal data. Experiments demonstrate that the single model trained based on FM-ViT can not only flexibly evaluate different modal samples, but also outperforms existing single-modal frameworks by a large margin.
  arXiv  Detail & Related papers  (2023-05-05T04:28:48Z)
• MA-ViT: Modality-Agnostic Vision Transformers for Face Anti-Spoofing [3.3031006227198003]
  We present Modality-Agnostic Vision Transformer (MA-ViT), which aims to improve the performance of arbitrary modal attacks with the help of multi-modal data. Specifically, MA-ViT adopts the early fusion to aggregate all the available training modalities data and enables flexible testing of any given modal samples. Experiments demonstrate that the single model trained on MA-ViT can not only flexibly evaluate different modal samples, but also outperforms existing single-modal frameworks by a large margin.
  arXiv  Detail & Related papers  (2023-04-15T13:03:44Z)
• Learning Visual Representation from Modality-Shared Contrastive Language-Image Pre-training [88.80694147730883]
  We investigate a variety of Modality-Shared Contrastive Language-Image Pre-training (MS-CLIP) frameworks. In studied conditions, we observe that a mostly unified encoder for vision and language signals outperforms all other variations that separate more parameters. Our approach outperforms vanilla CLIP by 1.6 points in linear probing on a collection of 24 downstream vision tasks.
  arXiv  Detail & Related papers  (2022-07-26T05:19:16Z)
• MIA-Former: Efficient and Robust Vision Transformers via Multi-grained Input-Adaptation [14.866949449862226]
  Vision Transformer (ViT) models are too computationally expensive to be fitted onto real-world resource-constrained devices. We propose a Multi-grained Input-adaptive Vision Transformer framework dubbed MIA-Former that can input-adaptively adjust the structure of ViTs. Experiments and ablation studies validate that the proposed MIA-Former framework can effectively allocate budgets adaptive to the difficulty of input images.
  arXiv  Detail & Related papers  (2021-12-21T22:06:24Z)
• Pruning Self-attentions into Convolutional Layers in Single Path [89.55361659622305]
  Vision Transformers (ViTs) have achieved impressive performance over various computer vision tasks. We propose Single-Path Vision Transformer pruning (SPViT) to efficiently and automatically compress the pre-trained ViTs. Our SPViT can trim 52.0% FLOPs for DeiT-B and get an impressive 0.6% top-1 accuracy gain simultaneously.
  arXiv  Detail & Related papers  (2021-11-23T11:35:54Z)
• Dynamic Clone Transformer for Efficient Convolutional Neural Netwoks [0.0]
  We introduce a novel concept termed multi-path fully connected pattern (MPFC) to rethink the interdependencies of topology pattern, accuracy and efficiency for ConvNets. Inspired by MPFC, we propose a dual-branch module named dynamic clone transformer (DCT) where one branch generates multiple replicas from inputs and another branch reforms those clones through a series of difference vectors conditional on inputs itself to produce more variants.
  arXiv  Detail & Related papers  (2021-06-12T13:42:28Z)

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.