# CropCop: An Auditable 120-Class Plant-Health Model from Benchmark Reconstruction to a Quantised Runtime Artifact

arXiv: https://arxiv.org/abs/2608.25539 · HTML: https://mlsanity.com/paper/2608.25539/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 21 of its 29 references that are in the map.
- First posted on arXiv 26 Aug 2026.

## Builds on (most cited first, 21 of 21)

- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [On Calibration of Modern Neural Networks](https://mlsanity.com/paper/1706.04599/) (2017)
- [A ConvNet for the 2020s](https://mlsanity.com/paper/2201.03545/) (2022)
- [Knowledge Distillation: A Survey](https://mlsanity.com/paper/2006.05525/) (2020)
- [Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference](https://mlsanity.com/paper/1712.05877/) (2017)
- [DINOv3](https://mlsanity.com/paper/2508.10104/) (2025)
- [Datasheets for Datasets](https://mlsanity.com/paper/1803.09010/) (2018)
- [Class-Balanced Loss Based on Effective Number of Samples](https://mlsanity.com/paper/1901.05555/) (2019)
- [Model Cards for Model Reporting](https://mlsanity.com/paper/1810.03993/) (2018)
- [Long-tail learning via logit adjustment](https://mlsanity.com/paper/2007.07314/) (2020)
- [Pervasive Label Errors in Test Sets Destabilize Machine Learning Benchmarks](https://mlsanity.com/paper/2103.14749/) (2021)
- [Using Deep Learning for Image-Based Plant Disease Detection](https://mlsanity.com/paper/1604.03169/) (2016)
- [An open access repository of images on plant health to enable the development of mobile disease diagnostics](https://mlsanity.com/paper/1511.08060/) (2015)
- [MobileNetV4 -- Universal Models for the Mobile Ecosystem](https://mlsanity.com/paper/2404.10518/) (2024)
- [PlantDoc: A Dataset for Visual Plant Disease Detection](https://mlsanity.com/paper/1911.10317/) (2019)
- [Do We Train on Test Data? Purging CIFAR of Near-Duplicates](https://mlsanity.com/paper/1902.00423/) (2019)
- [Benchmarking In-the-wild Multimodal Disease Recognition and A Versatile Baseline](https://mlsanity.com/paper/2408.03120/) (2024)
- [Intrinsic Self-Supervision for Data Quality Audits](https://mlsanity.com/paper/2305.17048/) (2023)
- [Mobile-Friendly Deep Learning for Plant Disease Detection: A Lightweight CNN Benchmark Across 101 Classes of 33 Crops](https://mlsanity.com/paper/2508.10817/) (2025)
- [Energy-Efficient Plant Monitoring via Knowledge Distillation](https://mlsanity.com/paper/2604.27178/) (2026)

## Abstract (arXiv)

A plant-health score can appear precise while resting on duplicated image families, a long-tailed label space, or a runtime file that was never evaluated. We present CropCop, a closed-set recognition system spanning 120 operational plant-health classes and an evidence chain from corpus reconstruction to direct execution of the final quantised artifact. Starting from 117,546 audited images, we rejected the inherited partition after confirming 3,233 duplicate relationships across split boundaries and froze a 109,107-image benchmark with zero crossings among the audited trusted leakage groups and a 151.7 largest-to-smallest class ratio. A fully fine-tuned DINOv3 ConvNeXt-Tiny reference achieved 98.51% accuracy and 96.87% macro-F1 on the locked internal test. A compact MobileNetV4 Conv-Medium derivative achieved 98.46% accuracy and 96.27% macro-F1 without being presented as evidence for a new distillation method. Validation-only post-training quantisation selected dynamic activations with per-channel weights, and the final 22.60 MiB ExecuTorch/XNNPACK PTE achieved 98.46% accuracy and 96.23% macro-F1 when executed directly. Only six of 16,363 top-1 decisions changed between the converted INT8 graph and the PTE, while paired analysis showed a modest class-balanced loss; an exploratory post hoc fruit-label slice localized a larger recall decline than aggregate accuracy revealed. CropCop establishes strong leakage-controlled internal recognition and software-runtime fidelity; it does not establish performance on unseen farms, camera pipelines, or physical Android hardware.