Sketch and Text Synergy: Fusing Structural Contours and Descriptive Attributes for Fine-Grained Image Retrieval

2026-04-17

AI Summary

This work addresses the performance bottleneck in fine-grained image retrieval caused by modality gaps between hand-drawn sketches and textual descriptions. To bridge this gap, the authors propose the STBIR framework, which for the first time systematically integrates the structural contours from sketches with the complementary color and texture cues from text. The approach enhances query robustness through curriculum learning, leverages category-level knowledge to guide feature space optimization, and introduces a multi-stage cross-modal alignment mechanism. The study also constructs the first fine-grained STBIR benchmark dataset, on which the proposed method significantly outperforms existing approaches, demonstrating the effectiveness and superiority of multimodal synergy in improving retrieval accuracy.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Graph Algorithms and Modeling for the Web: Querying, indexing, and retrieval in Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Fine-grained image retrieval via hand-drawn sketches or textual descriptions remains a critical challenge due to inherent modality gaps. While hand-drawn sketches capture complex structural contours, they lack color and texture, which text effectively provides despite omitting spatial contours. Motivated by the complementary nature of these modalities, we propose the Sketch and Text Based Image Retrieval (STBIR) framework. By synergizing the rich color and texture cues from text with the structural outlines provided by sketches, STBIR achieves superior fine-grained retrieval performance. First, a curriculum learning driven robustness enhancement module is proposed to enhance the model's robustness when handling queries of varying quality. Second, we introduce a category-knowledge-based feature space optimization module, thereby significantly boosting the model's representational power. Finally, we design a multi-stage cross-modal feature alignment mechanism to effectively mitigate the challenges of cross modal feature alignment. Furthermore, we curate the fine-grained STBIR benchmark dataset to rigorously validate the efficacy of our proposed framework and to provide data support as a reference for subsequent related research. Extensive experiments demonstrate that the proposed STBIR framework significantly outperforms state of the art methods.

Problem

• fine-grained image retrieval
• sketch-text fusion
• modality gap
• cross-modal retrieval
• structural contours

Innovation

• fine-grained image retrieval
• sketch-text fusion
• cross-modal alignment
• curriculum learning
• feature space optimization