Label-efficient seedling perception via semi-supervised learning and counterfactual spatiotemporal feature rectification
- 1School of Artificial Intelligence, Anhui Agricultural University, Hefei, 230036, China
- 2Institute of Intelligent Machines, HFIPS, Chinese Academy of Sciences, Hefei, 230031, China
- 3College of Electronics and Electrical Engineering, Anhui Agricultural University, Hefei, 230036, China
Abstract
In precision agriculture, assessing missing seedlings is essential for variable-rate reseeding and yield estimation. However, missing plants represent a semantic absence rather than explicit visual targets, making them difficult to define using conventional bounding-box annotations. To address this challenge, we propose a video-based, detection-driven gap inference framework for UAV wheat seedling phenotyping. Specifically, a semi-supervised Dual-Stream Anchor-Refinement Tracking (DART) mechanism is introduced to generate reliable pseudo-labels for visible seedlings from sparse temporal anchors, reducing dependence on dense manual annotations. To address visual fragmentation and weak target responses, a Masked Trajectory Inpainting Network (MTI-Net) combines object-aware masking with history-guided feature reconstruction to recover incomplete seedling representations using temporal context. The refined visible-seedling detections are then converted into agronomic gap indicators through a Dual-Scale Inertial Metric (DS-IM), which incorporates mechanical-sowing kinematic regularities and local growth-space constraints. Evaluations on the UAV-WheatSeedling dataset demonstrate positive mean gains when the proposed modules are embedded into RT-DETR and multiple YOLO-family backbones. When integrated with YOLO12-S, the method achieves 92.92 ± 0.17% [email protected], 69.33 ± 0.22% mAP@[0.5:0.95], and 86.59 ± 0.41% Recall. Under 50% sparse-anchor supervision, the proposed framework approaches the 100% sparse-anchor YOLO12 baseline in mAP@[0.5:0.95], while achieving higher mean Recall. For downstream gap inference, DS-IM achieves an F1-score of 84.53% and a frame-level count MAE of 0.229 gaps/frame. This study provides a data-efficient and interpretable framework for low-altitude UAV-based early seedling assessment.
Related
- Spatiotemporal Brain Patterns Elicited during Pain Perception in fMRI
- Counterfactual effect of Colombian protected areas on multispecies functional connectivity
- Intervention-Conditioned Counterfactual Human Digital Twin Generator
- scDEFT: A deep learning framework for drug-effect prediction and counterfactual reasoning
- An integrated explainable AI framework with counterfactual optimization for decision support
BibTeX
@article{nie2026labelefficient,
title = {{Label-efficient seedling perception via semi-supervised learning and counterfactual spatiotemporal feature rectification}},
author = {Yaoyao Nie and Huimin Ma and Xiaoshen Wang and Pengfei Zhang and Yi Wu and Xiaohong Li and Lichuan Gu},
journal = {Plant phenomics (Washington, D.C.)},
year = {2026},
month = sep,
doi = {10.1016/j.plaphe.2026.100278},
url = {https://europepmc.org/article/PMC/PMC13720541},
note = {CC BY},
}
RIS
TY - JOUR
TI - Label-efficient seedling perception via semi-supervised learning and counterfactual spatiotemporal feature rectification
AU - Yaoyao Nie
AU - Huimin Ma
AU - Xiaoshen Wang
AU - Pengfei Zhang
AU - Yi Wu
AU - Xiaohong Li
AU - Lichuan Gu
PY - 2026
DA - 2026/09/08
JO - Plant phenomics (Washington, D.C.)
DO - 10.1016/j.plaphe.2026.100278
UR - https://europepmc.org/article/PMC/PMC13720541
AB - In precision agriculture, assessing missing seedlings is essential for variable-rate reseeding and yield estimation. However, missing plants represent a semantic absence rather than explicit visual targets, making them difficult to define using conventional bounding-box annotations. To address this challenge, we propose a video-based, detection-driven gap inference framework for UAV wheat seedling phenotyping. Specifically, a semi-supervised Dual-Stream Anchor-Refinement Tracking (DART) mechanism is introduced to generate reliable pseudo-labels for visible seedlings from sparse temporal anchors, reducing dependence on dense manual annotations. To address visual fragmentation and weak target responses, a Masked Trajectory Inpainting Network (MTI-Net) combines object-aware masking with history-guided feature reconstruction to recover incomplete seedling representations using temporal context. The refined visible-seedling detections are then converted into agronomic gap indicators through a Dual-Scale Inertial Metric (DS-IM), which incorporates mechanical-sowing kinematic regularities and local growth-space constraints. Evaluations on the UAV-WheatSeedling dataset demonstrate positive mean gains when the proposed modules are embedded into RT-DETR and multiple YOLO-family backbones. When integrated with YOLO12-S, the method achieves 92.92 ± 0.17% [email protected], 69.33 ± 0.22% mAP@[0.5:0.95], and 86.59 ± 0.41% Recall. Under 50% sparse-anchor supervision, the proposed framework approaches the 100% sparse-anchor YOLO12 baseline in mAP@[0.5:0.95], while achieving higher mean Recall. For downstream gap inference, DS-IM achieves an F1-score of 84.53% and a frame-level count MAE of 0.229 gaps/frame. This study provides a data-efficient and interpretable framework for low-altitude UAV-based early seedling assessment.
SN - 2643-6515
ER -