Context-Aware Fine-Grained Product Recognition on Grocery Shelves

Product recognition is a fine-grained image retrieval problem because grocery stores can contain several thousand products on their shelves, some of which have minimal visual differences. Given that the products on the store’s shelves change frequently, it would be practical to build a sy...

Full description

Saved in:
Bibliographic Details
Main Authors: Lovre Antonio Budimir, Zoran Kalafatic, Marko Subasic, Sven Loncaric
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10849567/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1832576787472711680
author Lovre Antonio Budimir
Zoran Kalafatic
Marko Subasic
Sven Loncaric
author_facet Lovre Antonio Budimir
Zoran Kalafatic
Marko Subasic
Sven Loncaric
author_sort Lovre Antonio Budimir
collection DOAJ
description Product recognition is a fine-grained image retrieval problem because grocery stores can contain several thousand products on their shelves, some of which have minimal visual differences. Given that the products on the store’s shelves change frequently, it would be practical to build a system using a single reference image or a few per product. This task is very challenging in computer vision, whereas humans solve it more effortlessly by relying on contextual information. Our work incorporates semantic and spatial context into a novel product recognition method. We take advantage of stores being organized into different sections and aisles, where similar products are placed nearby. First, we propose the Hierarchical Auxiliary Loss (HAL) for learning an organized feature space in which products from the same category, usually placed in the same store section, are close to each other. Second, we propose the Context-Aware Query Expansion (CAQE) module for the inference phase, in which each feature vector of a product on a shelf is expanded with the feature vectors of neighboring products. The amount of information exchanged between the two products depends on the similarity of their feature vectors and spatial distance on the store shelves. To demonstrate the effectiveness of our contributions, we conducted detailed experiments on publicly available grocery product datasets and showed that our method achieves state-of-the-art results.
format Article
id doaj-art-d1caf5811e464c1686dff5af74f27a71
institution Kabale University
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-d1caf5811e464c1686dff5af74f27a712025-01-31T00:01:04ZengIEEEIEEE Access2169-35362025-01-0113168241683710.1109/ACCESS.2025.353274510849567Context-Aware Fine-Grained Product Recognition on Grocery ShelvesLovre Antonio Budimir0https://orcid.org/0000-0003-2146-7067Zoran Kalafatic1https://orcid.org/0000-0001-8918-9070Marko Subasic2https://orcid.org/0000-0002-4321-4557Sven Loncaric3https://orcid.org/0000-0002-4857-5351Faculty of Electrical Engineering and Computing, University of Zagreb, Zagreb, CroatiaFaculty of Electrical Engineering and Computing, University of Zagreb, Zagreb, CroatiaFaculty of Electrical Engineering and Computing, University of Zagreb, Zagreb, CroatiaFaculty of Electrical Engineering and Computing, University of Zagreb, Zagreb, CroatiaProduct recognition is a fine-grained image retrieval problem because grocery stores can contain several thousand products on their shelves, some of which have minimal visual differences. Given that the products on the store’s shelves change frequently, it would be practical to build a system using a single reference image or a few per product. This task is very challenging in computer vision, whereas humans solve it more effortlessly by relying on contextual information. Our work incorporates semantic and spatial context into a novel product recognition method. We take advantage of stores being organized into different sections and aisles, where similar products are placed nearby. First, we propose the Hierarchical Auxiliary Loss (HAL) for learning an organized feature space in which products from the same category, usually placed in the same store section, are close to each other. Second, we propose the Context-Aware Query Expansion (CAQE) module for the inference phase, in which each feature vector of a product on a shelf is expanded with the feature vectors of neighboring products. The amount of information exchanged between the two products depends on the similarity of their feature vectors and spatial distance on the store shelves. To demonstrate the effectiveness of our contributions, we conducted detailed experiments on publicly available grocery product datasets and showed that our method achieves state-of-the-art results.https://ieeexplore.ieee.org/document/10849567/Deep learningcontexthierarchical metric learningproduct recognitionquery expansion
spellingShingle Lovre Antonio Budimir
Zoran Kalafatic
Marko Subasic
Sven Loncaric
Context-Aware Fine-Grained Product Recognition on Grocery Shelves
IEEE Access
Deep learning
context
hierarchical metric learning
product recognition
query expansion
title Context-Aware Fine-Grained Product Recognition on Grocery Shelves
title_full Context-Aware Fine-Grained Product Recognition on Grocery Shelves
title_fullStr Context-Aware Fine-Grained Product Recognition on Grocery Shelves
title_full_unstemmed Context-Aware Fine-Grained Product Recognition on Grocery Shelves
title_short Context-Aware Fine-Grained Product Recognition on Grocery Shelves
title_sort context aware fine grained product recognition on grocery shelves
topic Deep learning
context
hierarchical metric learning
product recognition
query expansion
url https://ieeexplore.ieee.org/document/10849567/
work_keys_str_mv AT lovreantoniobudimir contextawarefinegrainedproductrecognitionongroceryshelves
AT zorankalafatic contextawarefinegrainedproductrecognitionongroceryshelves
AT markosubasic contextawarefinegrainedproductrecognitionongroceryshelves
AT svenloncaric contextawarefinegrainedproductrecognitionongroceryshelves