WE3DS: An RGB-D Image Dataset for Semantic Segmentation in Agriculture.

Smart farming (SF) applications rely on robust and accurate computer vision systems. An important computer vision task in agriculture is semantic segmentation, which aims to classify each pixel of an image and can be used for selective weed removal. State-of-the-art implementations use convolutional neural networks (CNN) that are trained on large image datasets. In agriculture, publicly available RGB image datasets are scarce and often lack detailed ground-truth information. In contrast to agriculture, other research areas feature RGB-D datasets that combine color (RGB) with additional distance (D) information. Such results show that including distance as an additional modality can improve model performance further. Therefore, we introduce WE3DS as the first RGB-D image dataset for multi-class plant species semantic segmentation in crop farming. It contains 2568 RGB-D images (color image and distance map) and corresponding hand-annotated ground-truth masks. Images were taken under natural light conditions using an RGB-D sensor consisting of two RGB cameras in a stereo setup. Further, we provide a benchmark for RGB-D semantic segmentation on the WE3DS dataset and compare it with a solely RGB-based model. Our trained models achieve up to 70.7% mean Intersection over Union (mIoU) for discriminating between soil, seven crop species, and ten weed species. Finally, our work confirms the finding that additional distance information improves segmentation quality.

DNA origami demonstrate the unique stimulatory power of single pMHCs as T cell antigens.

T cells detect with their T cell antigen receptors (TCRs) the presence of rare agonist peptide/MHC complexes (pMHCs) on the surface of antigen-presenting cells (APCs). How extracellular ligand binding triggers intracellular signaling is poorly understood, yet spatial antigen arrangement on the APC surface has been suggested to be a critical factor. To examine this, we engineered a biomimetic interface based on laterally mobile functionalized DNA origami platforms, which allow for nanoscale control over ligand distances without interfering with the cell-intrinsic dynamics of receptor clustering. When targeting TCRs via stably binding monovalent antibody fragments, we found the minimum signaling unit promoting efficient T cell activation to consist of two antibody-ligated TCRs within a distance of 20 nm. In contrast, transiently engaging antigenic pMHCs stimulated T cells robustly as well-isolated entities. These results identify pairs of antibody-bound TCRs as minimal receptor entities for effective TCR triggering yet validate the exceptional stimulatory potency of single isolated pMHC molecules.

A Multimodal Platform for Simultaneous T-Cell Imaging, Defined Activation, and Mechanobiological Characterization.

T-cell antigen recognition is accompanied by extensive morphological rearrangements of the contact zone between the T-cell and the antigen-presenting cell (APC). This process involves binding of the T-cell receptor (TCR) complex to antigenic peptides presented via MHC on the APC surface, the interaction of costimulatory and adhesion proteins, remodeling of the actin cytoskeleton, and the initiation of downstream signaling processes such as the release of intracellular calcium. However, multiparametric time-resolved analysis of these processes is hampered by the difficulty in recording the different readout modalities at high quality in parallel. In this study, we present a platform for simultaneous quantification of TCR distribution via total internal reflection fluorescence microscopy, of intracellular calcium levels, and of T-cell-exerted forces via atomic force microscopy (AFM). In our method, AFM cantilevers were used to bring single T-cells into contact with the activating surface. We designed the platform specifically to enable the study of T-cell triggering via functionalized fluid-supported lipid bilayers, which represent a widely accepted model system to stimulate T-cells in an antigen-specific manner. In this paper, we showcase the possibilities of this platform using primary transgenic T-cells triggered specifically via their cognate antigen presented by MHCII.

A micropatterning platform for quantifying interaction kinetics between the T cell receptor and an intracellular binding protein.

A complete understanding of signaling processes at the plasma membrane depends on a quantitative characterization of the interactions of the involved proteins. Fluorescence recovery after photobleaching (FRAP) is a widely used and convenient technique to obtain kinetic parameters on protein interactions in living cells. FRAP experiments to determine unbinding time constants for proteins at the plasma membrane, however, are often hampered by non-specific contributions to the fluorescence recovery signal. On the example of the interaction between the T cell receptor (TCR) and the Syk kinase ZAP70, we present here an approach based on protein micropatterning that allows the elimination of such non-specific contributions and considerably simplifies analysis of FRAP data. Specifically, detection and reference areas are created within single cells, each being either enriched or depleted in TCR, which permits the isolation of ZAP70-TCR binding in a straight-forward manner. We demonstrate the applicability of our method by comparing it to a conventional FRAP approach.