Track manipulation, locomotion and embodied systems.
Search papers, labs, and topics across Lattice.
This paper introduces UltraLight Luma, a novel, extremely compact encoder-decoder segmentation network with 13.1k parameters, designed for edge-deployable crop-row segmentation in agricultural robotics. By employing a progressive learning strategy and a specialized post-processing algorithm for navigation-line extraction, it achieves reliable performance with minimal computational and energy overhead. The network's efficiency is validated by its low inference energy consumption (25.23 mJ) and superior parameter efficiency compared to established architectures like U-Net and YOLO variants.
Agricultural robots can now achieve reliable crop-row segmentation with a network boasting 13.1k parameters and consuming just 25.23 mJ per inference, drastically reducing edge deployment constraints.
Reliable crop-row perception is essential for autonomous agricultural robots, but low-cost deployment is constrained by computation, memory, power, and inference time. This paper addresses crop-row detection and navigation-line extraction using UltraLight Luma, a compact encoder-decoder segmentation network with a progressive learning strategy and only 13.1k trainable parameters. The predicted crop-row masks are processed using a post-processing algorithm to estimate row alignment for downstream visual servoing. UltraLight Luma improves parameter efficiency over U-Net, YOLOv8, and YOLOv26 while maintaining reliable crop-row detection performance. The model required only 25.23 mJ per inference, demonstrating its suitability for resource-constrained agricultural robots.