U-shaped Multi-granularity Learning for Vision-Language Models

• URL: http://arxiv.org/abs/2607.14966v1
• Date: Thu, 16 Jul 2026 13:18:08 GMT
• Title: U-shaped Multi-granularity Learning for Vision-Language Models
• Abstract summary: We propose UPrompt, a multi-granularity prompt learning framework for vision-language models.<n>Similar to U-Net, UPrompt constructs parallel multi-granularity representations in both visual and textual modalities.<n>Extensive experiments on 17 benchmarks validate our effectiveness.
• Score: 8.898786482970456
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The prompt learning paradigm for vision-language models is effective yet faces a granularity dilemma: global prompts lack fine-grained semantic awareness, while local prompts ignore contextual associations, limiting cross-task generalization. This dilemma exists in dense prediction tasks. Inspired by U-Net, which unifies multi-level representations across granularities, we propose UPrompt, a U-shaped multi-granularity prompt learning framework for vision-language models. Similar to how U-Net integrates fine and coarse features through symmetric encoder-decoder pathways with cross-level connections, UPrompt constructs parallel multi-granularity representations in both visual and textual modalities, where coarse-to-fine cascaded enhancement propagates global context to refine local details, while fine-to-coarse hierarchical supervision ensures semantic consistency across scales. Extensive experiments on 17 benchmarks validate our effectiveness. UPrompt outperforms MAMET and VPKE by 4.1 and 7.3 rSum on MSCOCO, surpasses CoCoA-Mix by 5.09% in base-to-novel generalization, while maintaining competitive performance with minimal overhead (coarse-grained) and matching PSRC with 1/3 cost (medium-grained).

Related papers

• dinov3.seg: Open-Vocabulary Semantic Segmentation with DINOv3 [36.6036728217708]
  Open-Vocabulary Semantics (OVSS) assigns pixel-level labels from an open set of text-defined categories, demanding reliable generalization to unseen classes at inference.<n>We introduce dinov3.seg, extending dinov3.txt into a dedicated framework for OVSS.<n>Our contributions are four-fold. First, we design a task-specific architecture tailored to this backbone, systematically adapting established design principles from prior open-vocabulary segmentation work.<n>Second, we jointly leverage text embeddings aligned with both the global [] token and local patch-level visual features from ViT-based encoder
  arXiv  Detail & Related papers  (2026-03-19T23:57:28Z)
• LoGoSeg: Integrating Local and Global Features for Open-Vocabulary Semantic Segmentation [12.192429756057132]
  Open-vocabulary semantic segmentation (OVSS) extends traditional closed-set segmentation by enabling pixel-wise annotation for both seen and unseen categories.<n>LoGoSeg integrates three key innovations: (i) an object existence prior that dynamically weights relevant categories through global image-text similarity, effectively reducing hallucinations; (ii) a region-aware alignment module that establishes precise region-level visual-textual correspondences; and (iii) a dual-stream fusion mechanism that optimally combines local structural information with global semantic context.
  arXiv  Detail & Related papers  (2026-02-05T12:03:11Z)
• Multi-Perspective Subimage CLIP with Keyword Guidance for Remote Sensing Image-Text Retrieval [18.55080473948215]
  MPS-CLIP is a parameter-efficient framework designed to shift the retrieval paradigm from global matching to keyword-guided fine-grained alignment.<n>Experiments on the RSICD and RSITMD benchmarks demonstrate that MPS-CLIP achieves state-of-the-art performance with 35.18% and 48.40% mean Recall.
  arXiv  Detail & Related papers  (2026-01-26T06:16:53Z)
• Revisiting Multi-Task Visual Representation Learning [52.93947931352643]
  We introduce MTV, a principled multi-task visual pretraining framework.<n>We leverage high-capacity "expert" models to synthesize dense, structured pseudo-labels at scale.<n>Our results demonstrate that MTV achieves "best-of-both-worlds" performance.
  arXiv  Detail & Related papers  (2026-01-20T11:59:19Z)
• Topological Alignment of Shared Vision-Language Embedding Space [5.5522557994489246]
  ToMCLIP is a topology-aware framework aligning embedding spaces with topology-preserving constraints.<n>We show enhanced structural coherence of multilingual representations, higher zero-shot accuracy on the CIFAR-100, and stronger multilingual retrieval performance on the xFlickr&CO.
  arXiv  Detail & Related papers  (2025-10-13T01:36:38Z)
• Language-Driven Visual Consensus for Zero-Shot Semantic Segmentation [114.72734384299476]
  We propose a Language-Driven Visual Consensus (LDVC) approach, fostering improved alignment of semantic and visual information. We leverage class embeddings as anchors due to their discrete and abstract nature, steering vision features toward class embeddings. Our approach significantly boosts the capacity of segmentation models for unseen classes.
  arXiv  Detail & Related papers  (2024-03-13T11:23:55Z)
• UMG-CLIP: A Unified Multi-Granularity Vision Generalist for Open-World Understanding [90.74967596080982]
  This paper extends Contrastive Language-Image Pre-training (CLIP) with multi-granularity alignment. We develop a Unified Multi-Granularity learning framework, termed UMG-CLIP, which simultaneously empowers the model with versatile perception abilities. With parameter efficient tuning, UMG-CLIP surpasses current widely used CLIP variants and achieves state-of-the-art performance on diverse image understanding benchmarks.
  arXiv  Detail & Related papers  (2024-01-12T06:35:09Z)
• Towards Generalizable Referring Image Segmentation via Target Prompt and Visual Coherence [48.659338080020746]
  Referring image segmentation (RIS) aims to segment objects in an image conditioning on free-from text descriptions. We present a novel RIS approach, which substantially improves the generalization ability by addressing the two dilemmas mentioned above. Specially, to deal with unconstrained texts, we propose to boost a given expression with an explicit and crucial prompt, which complements the expression in a unified context.
  arXiv  Detail & Related papers  (2023-12-01T09:31:24Z)
• Global and Local Semantic Completion Learning for Vision-Language Pre-training [34.740507502215536]
  Cross-modal alignment plays a crucial role in vision-language pre-training models. We propose a novel Global and Local Semantic Completion Learning (GLSCL) task to facilitate global-local alignment and local-local alignment simultaneously.
  arXiv  Detail & Related papers  (2023-06-12T13:20:29Z)
• Robust Saliency-Aware Distillation for Few-shot Fine-grained Visual Recognition [57.08108545219043]
  Recognizing novel sub-categories with scarce samples is an essential and challenging research topic in computer vision. Existing literature addresses this challenge by employing local-based representation approaches. This article proposes a novel model, Robust Saliency-aware Distillation (RSaD), for few-shot fine-grained visual recognition.
  arXiv  Detail & Related papers  (2023-05-12T00:13:17Z)
• USER: Unified Semantic Enhancement with Momentum Contrast for Image-Text Retrieval [115.28586222748478]
  Image-Text Retrieval (ITR) aims at searching for the target instances that are semantically relevant to the given query from the other modality. Existing approaches typically suffer from two major limitations.
  arXiv  Detail & Related papers  (2023-01-17T12:42:58Z)
• MaPLe: Multi-modal Prompt Learning [54.96069171726668]
  We propose Multi-modal Prompt Learning (MaPLe) for both vision and language branches to improve alignment between the vision and language representations. Compared with the state-of-the-art method Co-CoOp, MaPLe exhibits favorable performance and achieves an absolute gain of 3.45% on novel classes.
  arXiv  Detail & Related papers  (2022-10-06T17:59:56Z)

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.