While I like to think that I've gotten pretty good at distinguishing among certain millets (and similar sized seeds), there are cases where I'm not so sure or have to look very closely. Sometimes this is at home, where I once almost mistook a jar of little millet for one of amaranth. From a distance, proso and foxtail millets are fairly obvious, but a jar with couscous (not even grain, of course) actually resembles the latter - until you get close. I've recently been experimenting with barnyard millet (likely Indian barnyard, imported from India) and browntop millet side-by-side in jars, they look quite similar. I'll return to this particular comparison another time, but the point is that reliably identifying small millets - out of the hull as I have them, or even in the hulls, can be a challenge even on an individual level. How much more complicated might this be if one had volumes of different millets in a storage or processing facility where they have to be maintained separately So it is of interest to note two recent articles discussing techniques for using AI / deep learning to automatically identify millet grains as well as plants: ----- Ravichandran, M., K. Jagan Mohan, and S. Sivasankaran, "A Comprehensive Review on Various Deep Learning Techniques for Identification and Classification of Millets," Proceedings of the International Conference on Artificial Intelligence and Secure Data Analytics (ICAISDA 2025), Atlantis Press, 2026. https://doi.org/10.2991/978-94-6239-616-6_38 Abstract "Millets, including pearl, finger, foxtail, proso, kodo, and little millet, are small-grained cereals known for their climate resilience. These crops require enhanced digital tools for species and cultivar recognition, disease detection, and origin and quality assessment. ML techniques based on neural networks, such as CNNs or combined models, have helped in millet classification from diverse data sources, including leaf and seed images, hyperspectral scans, and field-level imagery. This review elaborates on findings made between 2020 and 2025 while contrasting model families, datasets, and reported metrics across the modalities, pointing out key challenges such as diversity in datasets, environmental variability, explainability, and edge deployment." ------ Baisakh, B., P.K. Sethy, R. Gupta, et al, "A hybrid machine vision and handcrafted features fusion based approach for fine-grained millet classification," Scientific Repports 16, 2578 (2026). https://doi.org/10.1038/s41598-025-32311-4 Abstract: "Millets are increasingly recognized as climate-resilient, nutrient-rich grains essential for food and nutritional security. However, the limited availability of annotated datasets and the visual similarities among millet varieties such as Barnyard Millet, Little Millet, and Proso Millet pose significant challenges in their accurate identification and classification. To address these issues, this study proposes two novel hybrid frameworks that leverage both handcrafted and deep learning paradigms. A key component shared by both frameworks is the incorporation of a custom-designed handcrafted feature extraction process. The first framework introduces a Stacked Ensemble Machine Learning approach, that integrates handcrafted techniques and machine learning classifiers as Support Vector Machine (SVM), Random Forest (RF), and K-Nearest Neighbor (KNN) classifiers using a multinomial logistic regression meta-learner. The second framework adopts a bidirectional deep learning architecture; wherein handcrafted feature vectors are fused with the pretrained VGG19 model to achieve effective classification of different millet varieties. The stacked ensemble hybrid model achieved the 97% classification accuracy where as the second model achieved the highest accuracy of 96.7%, an F1-score of 0.967, and a ROC-AUC of 0.995, while maintaining excellent tradeoff between training and inference times." Don Osborn, PhD (East Lansing, MI, US) North American Millets Alliance