{"slug": "seeking-independent-reviewers-wheat-rust-ai-cross-dataset-generalisation", "title": "Seeking Independent Reviewers: Wheat Rust AI Cross-Dataset Generalisation", "summary": "Independent AI researcher Devanandha VS, an MSc Artificial Intelligence graduate of Ulster University, is seeking outside reviewers for Wheat Rust AI Detection, an open-source MobileNetV2-based classifier that sorts wheat leaves into Healthy, Brown Rust and Yellow Rust categories. The model reached 98.78% accuracy (728/737 correct predictions) on its internal validation set, but Devanandha reports that evaluations on independent datasets revealed substantial generalisation weaknesses, and he is asking collaborators to test cross-dataset performance, reproducibility, Grad-CAM explainability and model reliability via a dedicated GitHub issue. Devanandha states the project is \"an experimental research demonstration, not a production-ready agricultural diagnostic system.", "body_md": "Hello Hugging Face community!\n\nI’m Devanandha VS, an independent AI researcher and MSc Artificial Intelligence graduate from Ulster University, UK.\n\nI’m seeking independent researchers, computer-vision practitioners and open-source developers interested in evaluating an agricultural AI research project.\n\n**Wheat Rust AI Detection** is an open-source deep-learning project using a MobileNetV2-based classifier to identify three wheat leaf categories:\n\nHealthy\n\nBrown Rust\n\nYellow Rust\n\nThe classifier achieved **98.78% accuracy (728/737 correct predictions)** on its internal validation dataset.\n\nHowever, subsequent evaluations using independent datasets revealed substantial generalisation weaknesses.\n\nRather than presenting the initial accuracy as proof of real-world reliability, I investigated and publicly documented the model’s limitations, unsuccessful experiments and Grad-CAM explainability findings.\n\nThe project is an experimental research demonstration, not a production-ready agricultural diagnostic system.\n\nI welcome independent contributions in four areas:\n\n**1. Cross-Dataset Generalisation**\n\nEvaluate the model on independent wheat leaf datasets and investigate differences in performance across data sources.\n\n**2. Reproducibility**\n\nAttempt to reproduce the documented experiments, including preprocessing, evaluation methodology and dataset configurations.\n\n**3. Explainable AI**\n\nReview the Grad-CAM analysis and assess the usefulness and limitations of the generated explanations.\n\n**4. Model Reliability**\n\nExplore improved evaluation methods, relevant baselines, calibration techniques and strategies for addressing dataset shift.\n\nI’ve opened a dedicated GitHub issue for independent evaluation and collaboration:\n\n**Independent Reproduction and External Validation:**\n\n**Project Repository:**\n\nContributors are welcome to report findings, suggest improvements, submit pull requests or discuss potential methodological limitations.\n\nPositive feedback, critical assessments and unsuccessful reproduction attempts are equally welcome.\n\nWhen evaluating agricultural image classifiers across independent datasets, what factors would you investigate first?\n\nDifferences in image acquisition and environmental conditions?\n\nClass definitions and annotation quality?\n\nData preprocessing and dataset composition?\n\nDomain shift, class imbalance or model calibration?\n\nI’d appreciate hearing your experience and welcome collaboration on a more rigorous evaluation.\n\nThank you for your time and consideration!\n\n**Devanandha VS**\n\nMSc Artificial Intelligence, Ulster University\n\nIndependent AI Researcher", "url": "https://wpnews.pro/news/seeking-independent-reviewers-wheat-rust-ai-cross-dataset-generalisation", "canonical_source": "https://discuss.huggingface.co/t/seeking-independent-reviewers-wheat-rust-ai-cross-dataset-generalisation/191464#post_1", "published_at": "2026-11-05 14:04:46+00:00", "updated_at": "2026-11-05 14:23:17.273316+00:00", "lang": "en", "topics": ["computer-vision", "machine-learning", "ai-research"], "entities": ["Devanandha VS", "Ulster University", "Wheat Rust AI Detection", "MobileNetV2", "Hugging Face", "GitHub"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/seeking-independent-reviewers-wheat-rust-ai-cross-dataset-generalisation", "markdown": "https://wpnews.pro/news/seeking-independent-reviewers-wheat-rust-ai-cross-dataset-generalisation.md", "text": "https://wpnews.pro/news/seeking-independent-reviewers-wheat-rust-ai-cross-dataset-generalisation.txt", "jsonld": "https://wpnews.pro/news/seeking-independent-reviewers-wheat-rust-ai-cross-dataset-generalisation.jsonld"}}