One to All: Toward a Unified Model for Counting Cereal Crop Heads Based on Few-Shot Learning

Accurate counting of cereals crops, e.g., maize, rice, sorghum, and wheat, is crucial for estimating grain production and ensuring food security. However, existing methods for counting cereal crops focus predominantly on building models for specific crop head; thus, they lack generalizability to dif...

Full description

Saved in:
Bibliographic Details
Main Authors: Qiang Wang, Xijian Fan, Ziqing Zhuang, Tardi Tjahjadi, Shichao Jin, Honghua Huan, Qiaolin Ye
Format: Article
Language:English
Published: American Association for the Advancement of Science (AAAS) 2024-01-01
Series:Plant Phenomics
Online Access:https://spj.science.org/doi/10.34133/plantphenomics.0271
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850270017409515520
author Qiang Wang
Xijian Fan
Ziqing Zhuang
Tardi Tjahjadi
Shichao Jin
Honghua Huan
Qiaolin Ye
author_facet Qiang Wang
Xijian Fan
Ziqing Zhuang
Tardi Tjahjadi
Shichao Jin
Honghua Huan
Qiaolin Ye
author_sort Qiang Wang
collection DOAJ
description Accurate counting of cereals crops, e.g., maize, rice, sorghum, and wheat, is crucial for estimating grain production and ensuring food security. However, existing methods for counting cereal crops focus predominantly on building models for specific crop head; thus, they lack generalizability to different crop varieties. This paper presents Counting Heads of Cereal Crops Net (CHCNet), which is a unified model designed for counting multiple cereal crop heads by few-shot learning, which effectively reduces labeling costs. Specifically, a refined vision encoder is developed to enhance feature embedding, where a foundation model, namely, the segment anything model (SAM), is employed to emphasize the marked crop heads while mitigating complex background effects. Furthermore, a multiscale feature interaction module is proposed for integrating a similarity metric to facilitate automatic learning of crop-specific features across varying scales, which enhances the ability to describe crop heads of various sizes and shapes. The CHCNet model adopts a 2-stage training procedure. The initial stage focuses on latent feature mining to capture common feature representations of cereal crops. In the subsequent stage, inference is performed without additional training, by extracting domain-specific features of the target crop from selected exemplars to accomplish the counting task. In extensive experiments on 6 diverse crop datasets captured from ground cameras and drones, CHCNet substantially outperformed state-of-the-art counting methods in terms of cross-crop generalization ability, achieving mean absolute errors (MAEs) of 9.96 and 9.38 for maize, 13.94 for sorghum, 7.94 for rice, and 15.62 for mixed crops. A user-friendly interactive demo is available at http://cerealcropnet.com/, where researchers are invited to personally evaluate the proposed CHCNet. The source code for implementing CHCNet is available at https://github.com/Small-flyguy/CHCNet.
format Article
id doaj-art-af28bd640d4a445c97eec39a2b8e66d6
institution OA Journals
issn 2643-6515
language English
publishDate 2024-01-01
publisher American Association for the Advancement of Science (AAAS)
record_format Article
series Plant Phenomics
spelling doaj-art-af28bd640d4a445c97eec39a2b8e66d62025-08-20T01:52:49ZengAmerican Association for the Advancement of Science (AAAS)Plant Phenomics2643-65152024-01-01610.34133/plantphenomics.0271One to All: Toward a Unified Model for Counting Cereal Crop Heads Based on Few-Shot LearningQiang Wang0Xijian Fan1Ziqing Zhuang2Tardi Tjahjadi3Shichao Jin4Honghua Huan5Qiaolin Ye6Nanjing Forestry University, Nanjing 210037, China.Nanjing Forestry University, Nanjing 210037, China.Nanjing Forestry University, Nanjing 210037, China.University of Warwick, Coventry CV4 7AL, UK.Crop Phenomics Research Centre, Academy for Advanced Interdisciplinary Studies, Collaborative Innovation Centre for Modern Crop Production cosponsored by Province and Ministry, State Key Laboratory of Crop Genetics and Germplasm Enhancement, Nanjing Agricultural University, Nanjing 210095, China.Jiangsu Academy of Agricultural Sciences, Nanjing 210014, China.Nanjing Forestry University, Nanjing 210037, China.Accurate counting of cereals crops, e.g., maize, rice, sorghum, and wheat, is crucial for estimating grain production and ensuring food security. However, existing methods for counting cereal crops focus predominantly on building models for specific crop head; thus, they lack generalizability to different crop varieties. This paper presents Counting Heads of Cereal Crops Net (CHCNet), which is a unified model designed for counting multiple cereal crop heads by few-shot learning, which effectively reduces labeling costs. Specifically, a refined vision encoder is developed to enhance feature embedding, where a foundation model, namely, the segment anything model (SAM), is employed to emphasize the marked crop heads while mitigating complex background effects. Furthermore, a multiscale feature interaction module is proposed for integrating a similarity metric to facilitate automatic learning of crop-specific features across varying scales, which enhances the ability to describe crop heads of various sizes and shapes. The CHCNet model adopts a 2-stage training procedure. The initial stage focuses on latent feature mining to capture common feature representations of cereal crops. In the subsequent stage, inference is performed without additional training, by extracting domain-specific features of the target crop from selected exemplars to accomplish the counting task. In extensive experiments on 6 diverse crop datasets captured from ground cameras and drones, CHCNet substantially outperformed state-of-the-art counting methods in terms of cross-crop generalization ability, achieving mean absolute errors (MAEs) of 9.96 and 9.38 for maize, 13.94 for sorghum, 7.94 for rice, and 15.62 for mixed crops. A user-friendly interactive demo is available at http://cerealcropnet.com/, where researchers are invited to personally evaluate the proposed CHCNet. The source code for implementing CHCNet is available at https://github.com/Small-flyguy/CHCNet.https://spj.science.org/doi/10.34133/plantphenomics.0271
spellingShingle Qiang Wang
Xijian Fan
Ziqing Zhuang
Tardi Tjahjadi
Shichao Jin
Honghua Huan
Qiaolin Ye
One to All: Toward a Unified Model for Counting Cereal Crop Heads Based on Few-Shot Learning
Plant Phenomics
title One to All: Toward a Unified Model for Counting Cereal Crop Heads Based on Few-Shot Learning
title_full One to All: Toward a Unified Model for Counting Cereal Crop Heads Based on Few-Shot Learning
title_fullStr One to All: Toward a Unified Model for Counting Cereal Crop Heads Based on Few-Shot Learning
title_full_unstemmed One to All: Toward a Unified Model for Counting Cereal Crop Heads Based on Few-Shot Learning
title_short One to All: Toward a Unified Model for Counting Cereal Crop Heads Based on Few-Shot Learning
title_sort one to all toward a unified model for counting cereal crop heads based on few shot learning
url https://spj.science.org/doi/10.34133/plantphenomics.0271
work_keys_str_mv AT qiangwang onetoalltowardaunifiedmodelforcountingcerealcropheadsbasedonfewshotlearning
AT xijianfan onetoalltowardaunifiedmodelforcountingcerealcropheadsbasedonfewshotlearning
AT ziqingzhuang onetoalltowardaunifiedmodelforcountingcerealcropheadsbasedonfewshotlearning
AT tarditjahjadi onetoalltowardaunifiedmodelforcountingcerealcropheadsbasedonfewshotlearning
AT shichaojin onetoalltowardaunifiedmodelforcountingcerealcropheadsbasedonfewshotlearning
AT honghuahuan onetoalltowardaunifiedmodelforcountingcerealcropheadsbasedonfewshotlearning
AT qiaolinye onetoalltowardaunifiedmodelforcountingcerealcropheadsbasedonfewshotlearning