Overcoming field variability: unsupervised domain adaptation for enhanced crop-weed recognition in diverse farmlands
Recent developments in deep learning-based automatic weeding systems have shown promise for unmanned weed eradication. However, accurately distinguishing between crops and weeds in varying field conditions remains a challenge for these systems, as performance deteriorates when applied to new or diff...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2023-08-01
|
| Series: | Frontiers in Plant Science |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fpls.2023.1234616/full |
| _version_ | 1797745151956746240 |
|---|---|
| author | Talha Ilyas Talha Ilyas Jonghoon Lee Okjae Won Yongchae Jeong Yongchae Jeong Hyongsuk Kim Hyongsuk Kim |
| author_facet | Talha Ilyas Talha Ilyas Jonghoon Lee Okjae Won Yongchae Jeong Yongchae Jeong Hyongsuk Kim Hyongsuk Kim |
| author_sort | Talha Ilyas |
| collection | DOAJ |
| description | Recent developments in deep learning-based automatic weeding systems have shown promise for unmanned weed eradication. However, accurately distinguishing between crops and weeds in varying field conditions remains a challenge for these systems, as performance deteriorates when applied to new or different fields due to insignificant changes in low-level statistics and a significant gap between training and test data distributions. In this study, we propose an approach based on unsupervised domain adaptation to improve crop-weed recognition in new, unseen fields. Our system addresses this issue by learning to ignore insignificant changes in low-level statistics that cause a decline in performance when applied to new data. The proposed network includes a segmentation module that produces segmentation maps using labeled (training field) data while also minimizing entropy using unlabeled (test field) data simultaneously, and a discriminator module that maximizes the confusion between extracted features from the training and test farm samples. This module uses adversarial optimization to make the segmentation network invariant to changes in the field environment. We evaluated the proposed approach on four different unseen (test) fields and found consistent improvements in performance. These results suggest that the proposed approach can effectively handle changes in new field environments during real field inference. |
| first_indexed | 2024-03-12T15:19:23Z |
| format | Article |
| id | doaj.art-ec18781c143d4501a26903a7a488be3d |
| institution | Directory Open Access Journal |
| issn | 1664-462X |
| language | English |
| last_indexed | 2024-03-12T15:19:23Z |
| publishDate | 2023-08-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Plant Science |
| spelling | doaj.art-ec18781c143d4501a26903a7a488be3d2023-08-11T06:00:45ZengFrontiers Media S.A.Frontiers in Plant Science1664-462X2023-08-011410.3389/fpls.2023.12346161234616Overcoming field variability: unsupervised domain adaptation for enhanced crop-weed recognition in diverse farmlandsTalha Ilyas0Talha Ilyas1Jonghoon Lee2Okjae Won3Yongchae Jeong4Yongchae Jeong5Hyongsuk Kim6Hyongsuk Kim7Division of Electronics and Information Engineering, Jeonbuk National University, Jeonju-si, Republic of KoreaCore Research Institute of Intelligent Robots, Jeonbuk National University, Jeonju-si, Republic of KoreaCore Research Institute of Intelligent Robots, Jeonbuk National University, Jeonju-si, Republic of KoreaProduction Technology Research Division, National Institute of Crop Science, Rural Development Administration, Miryang, Republic of KoreaCore Research Institute of Intelligent Robots, Jeonbuk National University, Jeonju-si, Republic of KoreaDivision of Electronics Engineering, Jeonbuk National University, Jeonju-si, Republic of KoreaCore Research Institute of Intelligent Robots, Jeonbuk National University, Jeonju-si, Republic of KoreaDivision of Electronics Engineering, Jeonbuk National University, Jeonju-si, Republic of KoreaRecent developments in deep learning-based automatic weeding systems have shown promise for unmanned weed eradication. However, accurately distinguishing between crops and weeds in varying field conditions remains a challenge for these systems, as performance deteriorates when applied to new or different fields due to insignificant changes in low-level statistics and a significant gap between training and test data distributions. In this study, we propose an approach based on unsupervised domain adaptation to improve crop-weed recognition in new, unseen fields. Our system addresses this issue by learning to ignore insignificant changes in low-level statistics that cause a decline in performance when applied to new data. The proposed network includes a segmentation module that produces segmentation maps using labeled (training field) data while also minimizing entropy using unlabeled (test field) data simultaneously, and a discriminator module that maximizes the confusion between extracted features from the training and test farm samples. This module uses adversarial optimization to make the segmentation network invariant to changes in the field environment. We evaluated the proposed approach on four different unseen (test) fields and found consistent improvements in performance. These results suggest that the proposed approach can effectively handle changes in new field environments during real field inference.https://www.frontiersin.org/articles/10.3389/fpls.2023.1234616/fullcrop-weed recognitiondomain adaptationprecision agricultureartificial intelligencecrop phenotypingagricultural operations |
| spellingShingle | Talha Ilyas Talha Ilyas Jonghoon Lee Okjae Won Yongchae Jeong Yongchae Jeong Hyongsuk Kim Hyongsuk Kim Overcoming field variability: unsupervised domain adaptation for enhanced crop-weed recognition in diverse farmlands Frontiers in Plant Science crop-weed recognition domain adaptation precision agriculture artificial intelligence crop phenotyping agricultural operations |
| title | Overcoming field variability: unsupervised domain adaptation for enhanced crop-weed recognition in diverse farmlands |
| title_full | Overcoming field variability: unsupervised domain adaptation for enhanced crop-weed recognition in diverse farmlands |
| title_fullStr | Overcoming field variability: unsupervised domain adaptation for enhanced crop-weed recognition in diverse farmlands |
| title_full_unstemmed | Overcoming field variability: unsupervised domain adaptation for enhanced crop-weed recognition in diverse farmlands |
| title_short | Overcoming field variability: unsupervised domain adaptation for enhanced crop-weed recognition in diverse farmlands |
| title_sort | overcoming field variability unsupervised domain adaptation for enhanced crop weed recognition in diverse farmlands |
| topic | crop-weed recognition domain adaptation precision agriculture artificial intelligence crop phenotyping agricultural operations |
| url | https://www.frontiersin.org/articles/10.3389/fpls.2023.1234616/full |
| work_keys_str_mv | AT talhailyas overcomingfieldvariabilityunsuperviseddomainadaptationforenhancedcropweedrecognitionindiversefarmlands AT talhailyas overcomingfieldvariabilityunsuperviseddomainadaptationforenhancedcropweedrecognitionindiversefarmlands AT jonghoonlee overcomingfieldvariabilityunsuperviseddomainadaptationforenhancedcropweedrecognitionindiversefarmlands AT okjaewon overcomingfieldvariabilityunsuperviseddomainadaptationforenhancedcropweedrecognitionindiversefarmlands AT yongchaejeong overcomingfieldvariabilityunsuperviseddomainadaptationforenhancedcropweedrecognitionindiversefarmlands AT yongchaejeong overcomingfieldvariabilityunsuperviseddomainadaptationforenhancedcropweedrecognitionindiversefarmlands AT hyongsukkim overcomingfieldvariabilityunsuperviseddomainadaptationforenhancedcropweedrecognitionindiversefarmlands AT hyongsukkim overcomingfieldvariabilityunsuperviseddomainadaptationforenhancedcropweedrecognitionindiversefarmlands |