Compressing Semantic Segmentation Models for Agricultural IoT: A Systematic Study of Pruning Trade-Offs for Crop-Weed Segmentation.
- 1Department of Electrical Engineering and Information Technology (DIETI), University of Naples Federico II, 80125 Naples, Italy.
Abstract
Semantic segmentation of crops and weeds is an important component of selective weeding systems, but accurate segmentation networks can impose substantial computational and memory demands on resource-constrained agricultural platforms. This paper presents a systematic study of the accuracy-efficiency trade-offs involved in compressing semantic segmentation networks for agricultural edge computing. Six architectures spanning conventional CNN encoder-decoders, atrous and pyramid-pooling decoders, a mobile-oriented CNN, and two transformer scales are evaluated on the Crop/Weed Field Image Dataset (CWFID), with three random seeds per architecture. The most favourable compact model is subsequently evaluated under unstructured global magnitude pruning and dependency-aware structured channel pruning, using one-shot and iterative prune-and-fine-tune schedules at sparsity levels from 10% to 80%, with three seeds per configuration. Among the dense baselines, U-Net with a MobileNetV3-Large encoder matches the accuracy of the larger SegFormer-B2 model (88.85% versus 88.67% mIoU, a difference of 0.18% across three seeds, p = 0.58) with 3.7× fewer parameters than SegFormer-B2. Iterative unstructured pruning reaches 80% weight sparsity while retaining 89.04% mIoU; however, because the dense tensor dimensions are unchanged, this sparsity does not reduce the parameter count, GFLOPs, or model size on conventional hardware. Structured channel pruning, which does reduce all three, remains viable only up to 30% sparsity, reducing GFLOPs by 2.1× at a cost of 4.21 mIoU points and collapsing beyond 40%. Without fine-tuning, structured pruning causes severe degradation even at 10% channel sparsity. The results show that architecture selection has a major influence on the accuracy-efficiency trade-off, while the effects of pruning depend strongly on the pruning structure and fine-tuning strategy. Crop segmentation consistently degrades before the aggregate mIoU indicates severe failure, highlighting the need for class-wise evaluation when compressing crop-weed segmentation models.
Related
- WebXR-based mobile point cloud rendering and semantic segmentation using depth data and voxel fusion
- Deep learning for cardiac CT segmentation for congenital heart disease: A systematic review
- Visual rating versus volumetric segmentation of white matter hyperintensities for cognitive assessment: a memory clinic-based comparison and systematic review.
- Deep Learning for Real-Time Image Interpretation in Arthroscopy: A Systematic Review of Classification, Segmentation, and Detection Tasks
- Benchmark Readiness of Open-access Musculoskeletal Radiograph Datasets for Foundation Model Segmentation: A Systematic Audit
BibTeX
@article{amato2026compressing,
title = {{Compressing Semantic Segmentation Models for Agricultural IoT: A Systematic Study of Pruning Trade-Offs for Crop-Weed Segmentation.}},
author = {Flora Amato and Rajib Chandra Ghosh and Semanto Mondal},
journal = {Journal of imaging},
year = {2026},
month = oct,
doi = {10.3390/jimaging12100495},
url = {https://europepmc.org/article/MED/42930646},
note = {CC BY},
}
RIS
TY - JOUR
TI - Compressing Semantic Segmentation Models for Agricultural IoT: A Systematic Study of Pruning Trade-Offs for Crop-Weed Segmentation.
AU - Flora Amato
AU - Rajib Chandra Ghosh
AU - Semanto Mondal
PY - 2026
DA - 2026/10/08
JO - Journal of imaging
DO - 10.3390/jimaging12100495
UR - https://europepmc.org/article/MED/42930646
AB - Semantic segmentation of crops and weeds is an important component of selective weeding systems, but accurate segmentation networks can impose substantial computational and memory demands on resource-constrained agricultural platforms. This paper presents a systematic study of the accuracy-efficiency trade-offs involved in compressing semantic segmentation networks for agricultural edge computing. Six architectures spanning conventional CNN encoder-decoders, atrous and pyramid-pooling decoders, a mobile-oriented CNN, and two transformer scales are evaluated on the Crop/Weed Field Image Dataset (CWFID), with three random seeds per architecture. The most favourable compact model is subsequently evaluated under unstructured global magnitude pruning and dependency-aware structured channel pruning, using one-shot and iterative prune-and-fine-tune schedules at sparsity levels from 10% to 80%, with three seeds per configuration. Among the dense baselines, U-Net with a MobileNetV3-Large encoder matches the accuracy of the larger SegFormer-B2 model (88.85% versus 88.67% mIoU, a difference of 0.18% across three seeds, p = 0.58) with 3.7× fewer parameters than SegFormer-B2. Iterative unstructured pruning reaches 80% weight sparsity while retaining 89.04% mIoU; however, because the dense tensor dimensions are unchanged, this sparsity does not reduce the parameter count, GFLOPs, or model size on conventional hardware. Structured channel pruning, which does reduce all three, remains viable only up to 30% sparsity, reducing GFLOPs by 2.1× at a cost of 4.21 mIoU points and collapsing beyond 40%. Without fine-tuning, structured pruning causes severe degradation even at 10% channel sparsity. The results show that architecture selection has a major influence on the accuracy-efficiency trade-off, while the effects of pruning depend strongly on the pruning structure and fine-tuning strategy. Crop segmentation consistently degrades before the aggregate mIoU indicates severe failure, highlighting the need for class-wise evaluation when compressing crop-weed segmentation models.
SN - 2313-433X
ER -