Contrastive Bootstrap Representation Learning for Label-Efficient Agricultural Image Classification
Organizations: Department of Computer Science, Faculty of Graduate Studies for Statistical Research, Cairo University, Cairo, Egypt.
Abstract
Expert annotation remains a major bottleneck in agricultural image classification, where collecting images is often easier than obtaining reliable labels from domain specialists. Self-supervised learning can reduce this dependence by pretraining transferable representations on unlabelled imagery. Contrastive and bootstrap methods provide two established but distinct training signals, motivating an empirical study of a formulation that applies both signals within one representation path under severe downstream label scarcity. We present Contrastive Bootstrap Representation Learning (CBRL), which jointly optimises SimCLR-style NT-Xent contrastive learning and bidirectional BYOL-style exponential-moving-average target prediction within a shared online projection space. Using a ViT-L/16 backbone and 10% of the downstream labels, we report single-seed point estimates on four primary agricultural test benchmarks plus an exploratory 33-image New Plant Diseases evaluation. CBRL records the highest macro-F1 point estimate on the four primary benchmarks. On DeepWeeds, its 79.77% test macro-F1 is 32.29 points above the random-initialisation, label-only baseline and 15.78 points above the next-highest comparison point estimate; on PlantVillage, its 98.14% macro-F1 is 10.43 and 0.43 points higher, respectively. On PlantDoc, however, all methods fail to reach practically useful performance. These results identify the evaluated joint formulation as a promising candidate for label-limited transfer, but they do not establish objective synergy, statistical superiority, or robustness across stochastic runs.
Figures & tables
| Dataset | Classes | Labelled training examples |
|---|---|---|
| PlantVillage | 38 | 3,803 |
| New Plant Diseases † | 38 | 7,030 |
| DeepWeeds | 9 | 1,225 |
| PlantDoc | 28 | 235 |
| LUCAS Cover Photos | 59 | 13,669 |
| Dataset | Evaluations | Test |
|---|---|---|
| PlantVillage | 10 | 10 |
| New Plant Diseases † | 10 | 10 |
| DeepWeeds | 10 | 10 |
| PlantDoc | 10 | 10 |
| LUCAS | 10 | 10 |
| Total | 50 | 50 |
| Method | Acc. | F1 |
|---|---|---|
| Random-init ViT Dosovitskiy et al. (2021) | 90.89 | 87.71 |
| CBRL (ours) Chen et al. (2020) ; Grill et al. (2020) | 98.70 | 98.14 |
| ArcFace-SSL Deng et al. (2019) | 93.41 | 90.86 |
| HiFiC-SSL Mentzer et al. (2020) | 93.32 | 90.76 |
| DINOv2 Oquab et al. (2023) | 91.08 | 87.43 |
| BYOL Grill et al. (2020) | 94.91 | 92.70 |
Self-Supervised Visual Representation Learning
Self-supervised visual representation learning trains image, video, and 3D encoders on unlabeled data to produce transferable visual features.
See all 250 papers