FuseLIP: Multimodal Embeddings via Early Fusion of Discrete Tokens

• URL: http://arxiv.org/abs/2506.03096v1
• Date: Tue, 03 Jun 2025 17:27:12 GMT
• Title: FuseLIP: Multimodal Embeddings via Early Fusion of Discrete Tokens
• Authors: Christian Schlarmann, Francesco Croce, Nicolas Flammarion, Matthias Hein,
• Abstract summary: We present FuseLIP, an alternative architecture for multimodal embedding.<n>We propose a single transformer model which operates on an extended vocabulary of text and image tokens.<n>We show that FuseLIP outperforms other approaches in multimodal embedding tasks such as VQA and text-guided image transformation retrieval.
• Score: 56.752362642658504
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Contrastive language-image pre-training aligns the features of text-image pairs in a common latent space via distinct encoders for each modality. While this approach achieves impressive performance in several zero-shot tasks, it cannot natively handle multimodal inputs, i.e., encoding image and text into a single feature vector. As a remedy, it is common practice to use additional modules to merge the features extracted by the unimodal encoders. In this work, we present FuseLIP, an alternative architecture for multimodal embedding. Leveraging recent progress in discrete image tokenizers, we propose to use a single transformer model which operates on an extended vocabulary of text and image tokens. This early fusion approach allows the different modalities to interact at each depth of encoding and obtain richer representations compared to common late fusion. We collect new datasets for multimodal pre-training and evaluation, designing challenging tasks for multimodal encoder models. We show that FuseLIP outperforms other approaches in multimodal embedding tasks such as VQA and text-guided image transformation retrieval, while being comparable to baselines on unimodal tasks.

Related papers

• ToFu: Visual Tokens Reduction via Fusion for Multi-modal, Multi-patch, Multi-image Task [34.269081635534526]
  We propose ToFu, a visual encoder-agnostic, training-free Token Fusion strategy for high-resolution, multi-image, tasks.<n>We validate our approach on the well-established LLaVA-Interleave Bench, which covers challenging multi-image tasks.
  arXiv  Detail & Related papers  (2025-03-06T14:00:59Z)
• Towards Text-Image Interleaved Retrieval [49.96332254241075]
  We introduce the text-image interleaved retrieval (TIIR) task, where the query and document are interleaved text-image sequences.<n>We construct a TIIR benchmark based on naturally interleaved wikiHow tutorials, where a specific pipeline is designed to generate interleaved queries.<n>We propose a novel Matryoshka Multimodal Embedder (MME), which compresses the number of visual tokens at different granularity.
  arXiv  Detail & Related papers  (2025-02-18T12:00:47Z)
• MIFNet: Learning Modality-Invariant Features for Generalizable Multimodal Image Matching [54.740256498985026]
  Keypoint detection and description methods often struggle with multimodal data.<n>We propose a modality-invariant feature learning network (MIFNet) to compute modality-invariant features for keypoint descriptions in multimodal image matching.
  arXiv  Detail & Related papers  (2025-01-20T06:56:30Z)
• StitchFusion: Weaving Any Visual Modalities to Enhance Multimodal Semantic Segmentation [63.31007867379312]
  We propose StitchFusion, a framework that integrates large-scale pre-trained models directly as encoders and feature fusers.<n>We introduce a multi-directional adapter module (MultiAdapter) to enable cross-modal information transfer during encoding.<n>Our model achieves state-of-the-art performance on four multi-modal segmentation datasets with minimal additional parameters.
  arXiv  Detail & Related papers  (2024-08-02T15:41:16Z)
• Hierarchical Multi-modal Transformer for Cross-modal Long Document Classification [74.45521856327001]
  How to classify long documents with hierarchical structure texts and embedding images is a new problem. We propose a novel approach called Hierarchical Multi-modal Transformer (HMT) for cross-modal long document classification. Our approach uses a multi-modal transformer and a dynamic multi-scale multi-modal transformer to model the complex relationships between image features, and the section and sentence features.
  arXiv  Detail & Related papers  (2024-07-14T07:12:25Z)
• MM-Interleaved: Interleaved Image-Text Generative Modeling via Multi-modal Feature Synchronizer [106.79844459065828]
  This paper presents MM-Interleaved, an end-to-end generative model for interleaved image-text data. It introduces a multi-scale and multi-image feature synchronizer module, allowing direct access to fine-grained image features in the previous context. Experiments demonstrate the versatility of MM-Interleaved in recognizing visual details following multi-modal instructions and generating consistent images following both textual and visual conditions.
  arXiv  Detail & Related papers  (2024-01-18T18:50:16Z)
• Exchanging-based Multimodal Fusion with Transformer [19.398692598523454]
  We study the problem of multimodal fusion in this paper. Recent exchanging-based methods have been proposed for vision-vision fusion, which aim to exchange embeddings learned from one modality to the other. We propose a novel exchanging-based multimodal fusion model MuSE for text-vision fusion based on Transformer.
  arXiv  Detail & Related papers  (2023-09-05T12:48:25Z)
• Emu: Generative Pretraining in Multimodality [43.759593451544546]
  Transformer-based multimodal foundation model can seamlessly generate images and texts in multimodal context. Emu can serve as a generalist multimodal interface for both image-to-text and text-to-image tasks. Emu demonstrates superb performance compared to state-of-the-art large multimodal models.
  arXiv  Detail & Related papers  (2023-07-11T12:45:39Z)
• Attention-Based Multimodal Image Matching [16.335191345543063]
  We propose an attention-based approach for multimodal image patch matching using a Transformer encoder. Our encoder is shown to efficiently aggregate multiscale image embeddings while emphasizing task-specific appearance-invariant image cues. This is the first successful implementation of the Transformer encoder architecture to the multimodal image patch matching task.
  arXiv  Detail & Related papers  (2021-03-20T21:14:24Z)

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.