Adaptive Fine-Grained Sketch-Based Image Retrieval

Abstract

The recent focus on Fine-Grained Sketch-Based Image Re- trieval (FG-SBIR) has shifted towards generalising a model to new cat- egories without any training data from them. In real-world applications, however, a trained FG-SBIR model is often applied to both new cate- gories and different human sketchers, i.e., different drawing styles. Al- though this complicates the generalisation problem, fortunately, a hand- ful of examples are typically available, enabling the model to adapt to the new category/style. In this paper, we offer a novel perspective – instead of asking for a model that generalises, we advocate for one that quickly adapts, with just very few samples during testing (in a few-shot manner). To solve this new problem, we introduce a novel model-agnostic meta- learning (MAML) based framework with several key modifications: (1) As a retrieval task with a margin-based contrastive loss, we simplify the MAML training in the inner loop to make it more stable and tractable. (2) The margin in our contrastive loss is also meta-learned with the rest of the model. (3) Three additional regularisation losses are introduced in the outer loop, to make the meta-learned FG-SBIR model more effective for category/style adaptation. Extensive experiments on public datasets suggest a large gain over generalisation and zero-shot based approaches, and a few strong few-shot baselines.

Publication
European Conference on Computer Vision (ECCV) 2022
Animesh Gupta
Animesh Gupta

I am generally interested in deep learning algorithms for computer vision. I have been involved in the development of projects on Image classification, Sketch-Based Image Retrieval, Data-Efficient Training, and Semantic segmentation. I am currently working on training ViT models data efficiently and also on Sketch-Based Image Retrieval problems.