Learning Spatial Similarity Distribution for Few-shot Object Counting

TL;DR

This paper introduces a novel method for few-shot object counting that preserves and leverages the spatial distribution of similarity between query and exemplar images, leading to more accurate counts.

Contribution

It proposes a Spatial Similarity Distribution (SSD) network with a 4D similarity pyramid and new modules for enhanced feature matching and accurate density prediction.

Findings

Outperforms state-of-the-art on FSC-147 and CARPK datasets

Effectively captures complete similarity distribution information

Improves counting accuracy through feature cross enhancement

Abstract

Few-shot object counting aims to count the number of objects in a query image that belong to the same class as the given exemplar images. Existing methods compute the similarity between the query image and exemplars in the 2D spatial domain and perform regression to obtain the counting number. However, these methods overlook the rich information about the spatial distribution of similarity on the exemplar images, leading to significant impact on matching accuracy. To address this issue, we propose a network learning Spatial Similarity Distribution (SSD) for few-shot object counting, which preserves the spatial structure of exemplar features and calculates a 4D similarity pyramid point-to-point between the query features and exemplar features, capturing the complete distribution information for each point in the 4D similarity space. We propose a Similarity Learning Module (SLM) which applies the efficient center-pivot 4D convolutions on the similarity pyramid to map different similarity distributions to distinct predicted density values, thereby obtaining accurate count. Furthermore, we also introduce a Feature Cross Enhancement (FCE) module that enhances query and exemplar features mutually to improve the accuracy of feature matching. Our approach outperforms state-of-the-art methods on multiple datasets, including FSC-147 and CARPK. Code is available at https://github.com/CBalance/SSD.