Astrea: A MOE-based Visual Understanding Model with Progressive Alignment

• URL: http://arxiv.org/abs/2503.09445v2
• Date: Tue, 01 Apr 2025 03:10:50 GMT
• Title: Astrea: A MOE-based Visual Understanding Model with Progressive Alignment
• Authors: Xiaoda Yang, JunYu Lu, Hongshun Qiu, Sijing Li, Hao Li, Shengpeng Ji, Xudong Tang, Jiayang Xu, Jiaqi Duan, Ziyue Jiang, Cong Lin, Sihang Cai, Zejian Xie, Zhuoyang Song, Songxin Zhang,
• Abstract summary: Vision-Language Models (VLMs) based on Mixture-of-Experts (MoE) architectures have emerged as a pivotal paradigm in multimodal understanding.<n>We propose Astrea, a novel multi-expert collaborative VLM architecture based on progressive pre-alignment.
• Score: 10.943104653307294
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Vision-Language Models (VLMs) based on Mixture-of-Experts (MoE) architectures have emerged as a pivotal paradigm in multimodal understanding, offering a powerful framework for integrating visual and linguistic information. However, the increasing complexity and diversity of tasks present significant challenges in coordinating load balancing across heterogeneous visual experts, where optimizing one specialist's performance often compromises others' capabilities. To address task heterogeneity and expert load imbalance, we propose Astrea, a novel multi-expert collaborative VLM architecture based on progressive pre-alignment. Astrea introduces three key innovations: 1) A heterogeneous expert coordination mechanism that integrates four specialized models (detection, segmentation, classification, captioning) into a comprehensive expert matrix covering essential visual comprehension elements; 2) A dynamic knowledge fusion strategy featuring progressive pre-alignment to harmonize experts within the VLM latent space through contrastive learning, complemented by probabilistically activated stochastic residual connections to preserve knowledge continuity; 3) An enhanced optimization framework utilizing momentum contrastive learning for long-range dependency modeling and adaptive weight allocators for real-time expert contribution calibration. Extensive evaluations across 12 benchmark tasks spanning VQA, image captioning, and cross-modal retrieval demonstrate Astrea's superiority over state-of-the-art models, achieving an average performance gain of +4.7\%. This study provides the first empirical demonstration that progressive pre-alignment strategies enable VLMs to overcome task heterogeneity limitations, establishing new methodological foundations for developing general-purpose multimodal agents.

Related papers

• VL-Cogito: Progressive Curriculum Reinforcement Learning for Advanced Multimodal Reasoning [69.44871115752055]
  We propose an advanced multimodal reasoning model trained via a novel Progressive Curriculum Reinforcement Learning (PCuRL) framework.<n>PCuRL systematically guides the model through tasks of gradually increasing difficulty, substantially improving its reasoning abilities across diverse multimodal contexts.<n>The framework introduces two key innovations: (1) an online difficulty soft weighting mechanism, dynamically adjusting training difficulty across successive RL training stages; and (2) a dynamic length reward mechanism, which encourages the model to adaptively regulate its reasoning path length according to task complexity.
  arXiv  Detail & Related papers  (2025-07-30T12:23:21Z)
• MoCa: Modality-aware Continual Pre-training Makes Better Bidirectional Multimodal Embeddings [75.0617088717528]
  MoCa is a framework for transforming pre-trained VLM backbones into effective bidirectional embedding models.<n>MoCa consistently improves performance across MMEB and ViDoRe-v2 benchmarks, achieving new state-of-the-art results.
  arXiv  Detail & Related papers  (2025-06-29T06:41:00Z)
• REMoH: A Reflective Evolution of Multi-objective Heuristics approach via Large Language Models [39.85828629779943]
  Multi-objective optimization is fundamental in complex decision-making tasks.<n>Recent advances in Large Language Models (LLMs) offer enhanced explainability, adaptability, and reasoning.<n>This work proposes Reflective Evolution of Multi-objective Heuristics (REMoH), a novel framework integrating NSGA-II with LLM-based generation.
  arXiv  Detail & Related papers  (2025-06-09T13:38:28Z)
• Unveiling the Compositional Ability Gap in Vision-Language Reasoning Model [39.58344147240552]
  We investigate whether large vision-language models (VLMs) can compose capabilities across modalities or tasks under out-of-distribution conditions.<n>Our findings shed light on the current limitations of RL-based reasoning VLM training and provide actionable insights toward building models that reason compositionally across modalities and tasks.
  arXiv  Detail & Related papers  (2025-05-26T01:42:38Z)
• Co-AttenDWG: Co-Attentive Dimension-Wise Gating and Expert Fusion for Multi-Modal Offensive Content Detection [0.0]
  Multi-modal learning has emerged as a crucial research direction.<n>Existing approaches often suffer from insufficient cross-modal interactions and rigid fusion strategies.<n>We propose Co-AttenDWG, co-attention with dimension-wise gating, and expert fusion.<n>We show that Co-AttenDWG achieves state-of-the-art performance and superior cross-modal alignment.
  arXiv  Detail & Related papers  (2025-05-25T07:26:00Z)
• DeepPerception: Advancing R1-like Cognitive Visual Perception in MLLMs for Knowledge-Intensive Visual Grounding [61.26026947423187]
  Human experts excel at fine-grained visual discrimination by leveraging domain knowledge to refine perceptual features. Current Multimodal Large Language Models (MLLMs) struggle to integrate reasoning into visual perception. We propose DeepPerception, an MLLM enhanced with cognitive visual perception capabilities.
  arXiv  Detail & Related papers  (2025-03-17T04:06:34Z)
• Will Pre-Training Ever End? A First Step Toward Next-Generation Foundation MLLMs via Self-Improving Systematic Cognition [86.21199607040147]
  Self-Improving cognition (SIcog) is a self-learning framework for constructing next-generation foundation language models. We introduce Chain-of-Description, a step-by-step visual understanding method, and integrate structured chain-of-thought (CoT) reasoning to support in-depth multimodal reasoning. Extensive experiments demonstrate that SIcog produces next-generation foundation MLLMs with substantially improved multimodal cognition.
  arXiv  Detail & Related papers  (2025-03-16T00:25:13Z)
• SeFAR: Semi-supervised Fine-grained Action Recognition with Temporal Perturbation and Learning Stabilization [2.1682783789464968]
  Fine-grained Action Recognition (FAR) focuses on detailed semantic labels within shorter temporal duration.<n>Given the high costs of annotating labels and the substantial data needed for fine-tuning LLMs, we propose to adopt semi-supervised learning (SSL)<n>Our framework, SeFAR, incorporates several innovative designs to tackle these challenges.
  arXiv  Detail & Related papers  (2025-01-02T13:12:12Z)
• Chimera: Improving Generalist Model with Domain-Specific Experts [35.706585190958634]
  We introduce a scalable and low-cost multi-modal pipeline designed to boost the ability of existing LMMs with domain-specific experts.<n>Specifically, we design a progressive training strategy to integrate features from expert models into the input of a generalist LMM.<n>This results in a versatile model that excels across the chart, table, math, and document domains.
  arXiv  Detail & Related papers  (2024-12-08T16:10:42Z)
• MMEvol: Empowering Multimodal Large Language Models with Evol-Instruct [148.39859547619156]
  We propose MMEvol, a novel multimodal instruction data evolution framework.<n>MMEvol iteratively improves data quality through a refined combination of fine-grained perception, cognitive reasoning, and interaction evolution.<n>Our approach reaches state-of-the-art (SOTA) performance in nine tasks using significantly less data compared to state-of-the-art models.
  arXiv  Detail & Related papers  (2024-09-09T17:44:00Z)
• Alt-MoE:A Scalable Framework for Bidirectional Multimodal Alignment and Efficient Knowledge Integration [6.928469290518152]
  Multimodal learning has advanced significantly by aligning different modalities within shared latent spaces. Direct alignment struggles to fully leverage rich intra-modal knowledge, often requiring extensive training data to achieve cross-modal representation. We introduce Alt-MoE, a scalable multimodal alignment framework that employs a mixture of experts (MoE) model as a multi-directional connector across modalities.
  arXiv  Detail & Related papers  (2024-09-09T10:40:50Z)
• Asynchronous Multimodal Video Sequence Fusion via Learning Modality-Exclusive and -Agnostic Representations [19.731611716111566]
  We propose a Multimodal fusion approach for learning modality-Exclusive and modality-Agnostic representations. We introduce a predictive self-attention module to capture reliable context dynamics within modalities. A hierarchical cross-modal attention module is designed to explore valuable element correlations among modalities. A double-discriminator strategy is presented to ensure the production of distinct representations in an adversarial manner.
  arXiv  Detail & Related papers  (2024-07-06T04:36:48Z)
• Intuition-aware Mixture-of-Rank-1-Experts for Parameter Efficient Finetuning [50.73666458313015]
  Large Language Models (LLMs) have demonstrated significant potential in performing multiple tasks in multimedia applications. MoE has been emerged as a promising solution with its sparse architecture for effective task decoupling. Intuition-MoR1E achieves superior efficiency and 2.15% overall accuracy improvement across 14 public datasets.
  arXiv  Detail & Related papers  (2024-04-13T12:14:58Z)
• Unified Multi-modal Unsupervised Representation Learning for Skeleton-based Action Understanding [62.70450216120704]
  Unsupervised pre-training has shown great success in skeleton-based action understanding. We propose a Unified Multimodal Unsupervised Representation Learning framework, called UmURL. UmURL exploits an efficient early-fusion strategy to jointly encode the multi-modal features in a single-stream manner.
  arXiv  Detail & Related papers  (2023-11-06T13:56:57Z)
• Diversifying the Mixture-of-Experts Representation for Language Models with Orthogonal Optimizer [59.43462055143123]
  The Mixture of Experts (MoE) has emerged as a highly successful technique in deep learning. In this study, we shed light on the homogeneous representation problem, wherein experts in the MoE fail to specialize and lack diversity. We propose an alternating training strategy that encourages each expert to update in a direction to the subspace spanned by other experts.
  arXiv  Detail & Related papers  (2023-10-15T07:20:28Z)
• Exploiting Modality-Specific Features For Multi-Modal Manipulation Detection And Grounding [54.49214267905562]
  We construct a transformer-based framework for multi-modal manipulation detection and grounding tasks. Our framework simultaneously explores modality-specific features while preserving the capability for multi-modal alignment. We propose an implicit manipulation query (IMQ) that adaptively aggregates global contextual cues within each modality.
  arXiv  Detail & Related papers  (2023-09-22T06:55:41Z)

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.