Learning Cross-Modal Deep Embeddings for Multi-Object Image Retrieval using Text and Sketch

2018/04/28 by Sounak Dey, Anjan Dutta, Dey, Sounak +9
Computer Science · #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #cs.CV

paper · pdf · doi:10.48550/arxiv.1804.10819

Accepted at ICPR 2018

arxiv created 2018/04/28 · arxiv updated 2018/05/01

Abstract

In this work we introduce a cross modal image retrieval system that allows both text and sketch as input modalities for the query. A cross-modal deep network architecture is formulated to jointly model the sketch and text input modalities as well as the the image output modality, learning a common embedding between text and images and between sketches and images. In addition, an attention model is used to selectively focus the attention on the different objects of the image, allowing for retrieval with multiple objects in the query. Experiments show that the proposed method performs the best in both single and multiple object image retrieval in standard datasets.

Related

• Cross-Modal Fusion Distillation for Fine-Grained Sketch-Based Image Retrieval
• Revisiting Cross Modal Retrieval
• Cross-Modal Attention Alignment Network with Auxiliary Text Description for zero-shot sketch-based image retrieval
• Do Cross Modal Systems Leverage Semantic Relationships?
• Look, Imagine and Match: Improving Textual-Visual Cross-Modal Retrieval with Generative Models