Bounding overwatch

About: Bounding overwatch is a research topic. Over the lifetime, 966 publications have been published within this topic receiving 15156 citations.

TasselNetV3: Explainable Plant Counting With Guided Upsampling and Background Suppression

Abstract: Fast and accurate plant counting tools affect revolution in modern agriculture. Agricultural practitioners, however, expect the output of the tools to be not only accurate but also explainable. Such explainability often refers to the ability to infer which instance is counted. One intuitive way is to generate a bounding box for each instance. Nevertheless, compared with counting by detection, plant counts can be inferred more directly in the local count framework, while one thing reproaching this paradigm is its poor explainability of output visualization. In particular, we find that the poor explainability becomes a bottleneck limiting the counting performance. To address this, we explore the idea of guided upsampling and background suppression where a novel upsampling operator is proposed to allow count redistribution, and segmentation decoders with different fusion strategies are investigated to suppress background, respectively. By integrating them into our previous counting model TasselNetV2, we introduce TasselNetV3 series: TasselNetV3-Lite and TasselNetV3-Seg. We validate the TasselNetV3 series on three public plant counting data sets and a new unmanned aircraft vehicle (UAV)-based data set, covering maize tassels counting, wheat ears counting, and rice plants counting. Extensive results show that guided upsampling and background suppression not only improve counting performance but also enable explainable visualization. Aside from state-of-the-art performance, we have several interesting observations: 1) a limited-receptive-field counter in most cases outperforms a large-receptive-field one; 2) it is sufficient to generate empirical segmentation masks from dotted annotations; 3) middle fusion is a good choice to integrate foreground-background a priori knowledge; and 4) decoupling the learning of counting and segmentation matters.

Strategies for Bounding Volume Hierarchy Updates for Ray Tracing of Deformable Models

Abstract: In this paper, we describe strategies for bounding volume hierarchy updates for ray tracing of deformable models. By using pre-built hierarchy structures and a lazy evaluation technique for updating the bounding volumes, the hierarchy reconstruction can be made very ecien tly. Experiments show that for deforming triangle meshes the reconstruction time of the bounding volume hierarchies per frame can be reduced by an order of magnitude compared to previous approaches, which also results in a signican t speed-up in the total rendering time for many types of dynamically changing scenes. We believe our approach is a step towards interactive ray tracing of scenes where moving objects can be dynamically changed in non-deterministic ways.

Quadrupedal running with a flexible torso: control and speed transitions with sums-of-squares verification

Abstract: This paper studies quadrupedal bounding in the presence of flexible torso and compliant legs with non-trivial inertia, and it proposes a method for speed transitions by sequentially composing locally stable bounding motions corresponding to different speeds. First, periodic bounding motions are generated simply by positioning the legs during flight via suitable (virtual) holonomic constraints imposed on the leg angles; at this stage, no control effort is developed on the support legs, producing efficient, nearly passive, bounding gaits. The resulting motions are then stabilized by a hybrid control law which coordinates the movement of the torso and legs in continuous time, and updates the leg touchdown angles in an event-based fashion. Finally, through sums-of-squares programming, formally verified estimates of the domain of attraction of stable fixed points are employed to realize stable speed transitions by switching among different bounding gaits in a sequential fashion.