2016/11/28 by Ang Li, Jin Sun, Li, Ang +9 · 1 citation
Computer Science · Mathematics · #Advanced Image and Video Retrieval Techniques #Artificial intelligence #Bounding overwatch #Computational Geometry (cs.CG) #Computer Vision and Pattern Recognition (cs.CV) #Computer science #Computer vision #Data mining #FOS: Computer and information sciences #Image (mathematics) #Image Retrieval and Classification Techniques #Inference #Information Retrieval (cs.IR) #Information retrieval #Matching (statistics) #Mathematics #Multimodal Machine Learning Applications #Object (grammar) #Pairwise comparison #Pattern recognition (psychology) #Perspective (graphical) #Relation (database) #Set (abstract data type) #Spatial relation #cs.CG #cs.CV #cs.IR
paper · pdf · doi:10.48550/arxiv.1611.09392
published in arXiv (Cornell University) (Cornell University) · CVPR 2017
openalex publication_date 2016/11/28 · arxiv created 2017/04/11 · arxiv updated 2017/04/13 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05
Spatial relationships between objects provide important information for text-based image retrieval. As users are more likely to describe a scene from a real world perspective, using 3D spatial relationships rather than 2D relationships that assume a particular viewing direction, one of the main challenges is to infer the 3D structure that bridges images with users' text descriptions. However, direct inference of 3D structure from images requires learning from large scale annotated data. Since interactions between objects can be reduced to a limited set of atomic spatial relations in 3D, we study the possibility of inferring 3D structure from a text description rather than an image, applying physical relation models to synthesize holistic 3D abstract object layouts satisfying the spatial constraints present in a textual description. We present a generic framework for retrieving images from a textual description of a scene by matching images with these generated abstract object layouts. Images are ranked by matching object detection outputs (bounding boxes) to 2D layout candidates (also represented by bounding boxes) which are obtained by projecting the 3D scenes with sampled camera directions. We validate our approach using public indoor scene datasets and show that our method outperforms baselines built upon object occurrence histograms and learned 2D pairwise relations.