Composite Sketch+Text Queries for Retrieving Objects with Elusive Names and Complex Interactions

2024-03-24

AAAI Conference on Artificial Intelligence

AI Summary

To address the challenge of retrieving complex interactive objects (e.g., “a numbat digging on the ground”) that are difficult for non-native speakers or vocabulary-limited users to name yet sketchable—or describable yet hard to sketch, this paper introduces Composite Sketch-Text-Based Image Retrieval (CSTBIR), a novel cross-modal retrieval task. We present the first large-scale CSTBIR benchmark comprising 2 million composite queries and 108K images. To tackle this task, we propose STNet, a multi-modal Transformer architecture optimized via multi-objective learning: it jointly models sketch–image local alignment, text–image joint embedding, and hierarchical losses for object localization, cross-modal matching, and interaction modeling. Extensive experiments demonstrate that STNet significantly outperforms state-of-the-art text-based (TBIR), sketch-based (SBIR), and hybrid retrieval methods on composite queries. Both code and dataset are publicly released to advance fine-grained cross-modal understanding research.

Technology Category

• Computer Vision: Image and Video Retrieval
• Machine Learning: Multimodal Learning
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Querying, indexing, and retrieval in Web-related graphs
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data

Abstract

Non-native speakers with limited vocabulary often struggle to name specific objects despite being able to visualize them, e.g., people outside Australia searching for ‘numbats.’ Further, users may want to search for such elusive objects with difficult-to-sketch interactions, e.g., “numbat digging in the ground.” In such common but complex situations, users desire a search interface that accepts composite multimodal queries comprising hand-drawn sketches of “difficult-to-name but easy-to-draw” objects and text describing “difficult-to-sketch but easy-to-verbalize” object's attributes or interaction with the scene. This novel problem statement distinctly differs from the previously well-researched TBIR (text-based image retrieval) and SBIR (sketch-based image retrieval) problems. To study this under-explored task, we curate a dataset, CSTBIR (Composite Sketch+Text Based Image Retrieval), consisting of ~2M queries and 108K natural scene images. Further, as a solution to this problem, we propose a pretrained multimodal transformer-based baseline, STNet (Sketch+Text Network), that uses a hand-drawn sketch to localize relevant objects in the natural scene image, and encodes the text and image to perform image retrieval. In addition to contrastive learning, we propose multiple training objectives that improve the performance of our model. Extensive experiments show that our proposed method outperforms several state-of-the-art retrieval methods for text-only, sketch-only, and composite query modalities. We make the dataset and code available at: https://vl2g.github.io/projects/cstbir.

Problem

• Develops a multimodal search interface for elusive objects.
• Combines sketches and text for complex object retrieval.
• Introduces CSTBIR dataset and STNET model for retrieval.

Innovation

• Multimodal composite query interface
• Pretrained transformer for localization
• Contrastive learning with multiple objectives