This study evaluates the instance‑segmentation performance of YOLOv11 and YOLOv8 on immature green apples in orchard settings. YOLO11n‑seg achieved the highest mask precision (0.831), while YOLO11m‑seg and YOLO11l‑seg excelled in non‑occluded and occluded fruitlet segmentation. YOLOv8n, however, outperformed the YOLO11 series in inference speed, reaching 3.3 ms compared to 4.8 ms for the fastest YOLO11 model.
By Ranjan Sapkota, Manoj Karkee
The paper introduces the Selective Spatio-Temporal Fruit Inspector (S$^2$FI), a low‑cost, high‑efficiency fruit sorting system that uses a 3D convolutional encoder and a Mamba‑based backbone to extract spatiotemporal features and adaptively weight informative frames during fruit rotation. The hardware design features a dual‑conveyor and curved rail system powered by household electricity, enabling fast deployment and lightweight embedded operation. Experiments on a newly built dataset of 6,460 samples demonstrate that S$^2$FI achieves the best balance of accuracy, speed, and practicality compared to baseline methods.
By Hao Tian, Jie Hu, Guoqi Shan, Fan Wang, Wenzhuo Zhang, Youshan Zhang
Albireo is an adaptive, energy‑efficient inference framework for video object detection on edge devices that wraps existing detectors without modification. It uses a 10‑dimensional Kalman filter per active object to decide when to skip detector calls, predicting bounding boxes on skipped frames at near‑zero GPU cost. Evaluated on BDD100K with YOLO and RF‑DETR detectors on NVIDIA Jetson AGX Thor and Orin, Albireo maintains AP@50 within ±1.2 pp of full‑frame inference while reducing energy consumption by 12.1–17.6 % and improving accuracy for some models.
By Amir Taherin, Jos\'e Cano, Bin Ren, Yanzhi Wang, David Kaeli
The paper introduces a lightweight multimodal vision‑language framework based on TinyCLIP for fine‑grained classification of early‑stage apple fruitlet anatomy (calyx, fruitlet body, peduncle) in orchard images. Using a dataset of 600 high‑resolution RGB images, the model employs domain‑specific language prompts and a sliding‑window inference strategy to produce interpretable heatmaps for whole‑image localization. Achieving macro‑F1 of 0.93 on an NVIDIA T4 GPU and maintaining accuracy after INT8 quantization, the system is optimized for edge deployment on NVIDIA Jetson hardware with model sizes around 127‑137 MB and millisecond‑level inference.
By Ranjan Sapkota, William Bu, Chen Chen, Yunjun Xu, Manoj Karkee
arXiv:2608.23636v1 Announce Type: new
Abstract: Small-object detection and instance segmentation remain challenging in orchard environments because of green-on-green similarity, occlusion, and limite...
By Ranjan Sapkota, Manoj Karkee
RootDown presents an end‑to‑end image‑based pipeline for detecting partially occluded sugar beets on loader conveyors, a key step toward automated quality inspection. The authors introduce a dedicated benchmark and an automatic annotation strategy that uses the SAM 3 foundation model to generate bounding boxes, comparing these labels with manually annotated ones. They also evaluate deployment on edge hardware, showing that a Raspberry Pi 5 with a Hailo accelerator can perform high‑performance, low‑power inference suitable for mobile outdoor use.
By Gerardus Croonen, Andreas Trondl, Julia Simon, Daniel Steininger
The paper evaluates the new YOLO26 architecture, which offers NMS-free end-to-end inference and is tailored for CPU-based edge devices, against three earlier Ultralytics models (YOLOv5u, YOLOv8, and YOLO11) in aquaculture fish mortality detection. Across nano, small, and medium scales, all models achieved similar detection accuracy on a full dataset, but differences emerged in data efficiency and deployment performance: YOLOv8 reached 90% mAP50 with only 400 images, while YOLO26 variants needed 1,000 images; YOLO26n was fastest on a Raspberry Pi 5 (7.51 FPS), whereas YOLOv5mu led on CPU-based hardware. The study concludes that architectural novelty alone does not dictate suitability for edge AI in aquaculture; training data size, target hardware, and inference needs must be jointly considered.
By Rakesh Ranjan, Gajanan S. Kothawade, Kata Sharrer, Scott Tsukuda, Christopher Good
Commercial greenhouse cucumber production is graded by fruit length, which drives harvest scheduling, labour allocation, and logistics. Manual measurement with thread or caliper is accurate but infeasible at commercial scale.
The paper compares Ultralytics YOLO27, YOLO26, YOLO11, and YOLOv8 for detecting and segmenting small fruit parts in orchard settings. It evaluates five model scales across 30 experiments, finding that YOLO11s-960 and YOLO26s-960 achieve the best mask and box mAP scores while maintaining efficient parameter counts. The study also highlights the difficulty of peduncle detection and provides publicly available code and models for reproducibility.
By Ranjan Sapkota, Manoj Karkee
The paper introduces a geometry‑aware post‑detection framework that resolves overlapping plant instances in RGB UAV imagery by combining object detection with geometric clustering of plant components. It uses component centroids and radial intersection points (RIPs) to determine whether a detected region contains one or two plants, applying K‑means or Gaussian mixture models and density filtering to improve clustering accuracy. Evaluated on eggplant and tomato crops, the method achieved high F1‑scores (0.89 for eggplant, 0.75 for tomato) and demonstrated that geometric reasoning can enhance plant‑level interpretation without requiring additional sensors or retraining.
By Ik Jae Lee, Hieu D. Nguyen, Mahbubur Meenar, Carlos Morrison Martinez, Cameron Connelly
The paper introduces Fog-based RetailLightNet (F RLN), an ultra‑lightweight deep learning model designed for real‑time video analytics in smart retail environments. By deploying the model on edge devices such as the NVIDIA Jetson Nano and using motion‑gated frame selection, the system reduces latency and computational load while maintaining high precision (0.988), recall (0.976), and F1‑score (0.976). The approach leverages publicly available retail datasets (SKU110K and RPC) and depth‑wise separable convolutions to enable efficient shelf monitoring and queue assessment.
By Mayank Atreya, Navin Chhibber, Shilpa Ajay, R Loshini, K Anitha
LeafTrackNet is a deep learning framework that combines a YOLOv10-based leaf detector with a MobileNetV3-based embedding network to track individual leaves over time. The authors introduce CanolaTrack, a large benchmark dataset of 5,704 RGB images with 31,840 annotated leaf instances from 184 canola plants. When evaluated without prior fine‑tuning, LeafTrackNet outperforms existing methods on CanolaTrack, KOMATSUNA, and MSU‑PID datasets, achieving HOTA scores of 88.03, 87.33, and 74.20 respectively.
By Shanghua Liu, Majharulislam Babor, Christoph Verduyn, Breght Vandenberghe, Bruno Betoni Parodi, Cornelia Weltzien, Marina M. -C. H\"ohne