Search

International audience; Traditional Convolutional Neural Networks (CNN) for semantic segmentation of images use 2D convolution operations. While the spatial inductive bias of 2D convolutions allow CNNs to build hierarchical feature representations, they require that the whole feature maps are kept in memory until the end of the inference. This is not ideal for memory and latency-critical applications such as real-time on-board satellite image segmentation. In this paper, we propose a new neural network architecture for semantic segmentation, "Scan-nerNet", based on a Recurrent 1D Convolutional architecture. Our network performs a segmentation of the input image lineby-line, and thus reduces the memory footprint and output latency. These characteristics make it ideal for on-the-fly segmentation of images on-board satellites equipped with push broom sensors such as Landsat 8, or satellites with limited compute capabilities, such as Cubesats. We perform cloud segmentation experiments on embedded hardware and show that our method offers a good compromise between accuracy, memory usage and latency.

International audience; We present an algorithm for detecting planar primitives from unorganized 3D point clouds. Departing from an initial configuration, the algorithm refines both the continuous plane parameters and the discrete assignment of input points to them by seeking high fidelity, high simplicity and high completeness. Our key contribution relies upon the design of an exploration mechanism guided by a multiobjective energy function. The transitions within the large solution space are handled by five geometric operators that create, remove and modify primitives. We demonstrate the potential of our method on a variety of scenes, from organic shapes to man-made objects, and sensors, from multiview stereo to laser. We show its efficacy with respect to existing primitive fitting approaches and illustrate its applicative interest in compact mesh reconstruction, when combined with a plane assembly method.

International audience; Automatic road graph extraction from aerial and satellite images is a long-standing challenge. Existing algorithms are either based on pixel-level segmentation followed by vectorization, or on iterative graph construction using next move prediction. Both of these strategies suffer from severe drawbacks, in particular high computing resources and incomplete outputs. By contrast, we propose a method that directly infers the final road graph in a single pass. The key idea consists in combining a Fully Convolutional Network in charge of locating points of interest such as intersections, dead ends and turns, and a Graph Neural Network which predicts links between these points. Such a strategy is more efficient than iterative methods and allows us to streamline the training process by removing the need for generation of starting locations while keeping the training end-to-end. We evaluate our method against existing works on the popular RoadTracer dataset and achieve competitive results. We also benchmark the speed of our method and show that it outperforms existing approaches. Our method opens the possibility of in-flight processing on embedded devices for applications such as real-time road network monitoring and alerts for disaster response.

Publication in the conference proceedings of EUSIPCO, Poznan, Poland, 2007

Visual Relocalization is a key technology in many computer vision applications. Traditional visual relocalization is mainly achieved through geometric methods, while PoseNet introduces convolutional neural network in visual relocalization for the first time to realize real-time camera pose estimation based on a single image. Aiming at the problem of accuracy and robustness of the current PoseNet algorithm in complex environment, this paper proposes and implements a new high-precision robust camera pose calculation method (LRF-PoseNet). This method directly adjusts the size of the input image without cropping, so as to increase the receptive field of the training image. Then, the image and the corresponding pose tags are input into the improved LSTM-based PoseNet network for training, and the Adam optimizer is used to optimize the network. Finally, the trained network is used to estimate the camera pose. Experimental results on open RGB dataset show that the proposed method in this paper can obtain more accurate camera pose compared with the existing CNN-based methods., International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLIII-B2-2021, ISSN:1682-1750, ISSN:2194-9034, ISSN:1682-1777

International audience; We present an algorithm for extracting and vectorizing objects in images with polygons. Departing from a polygonal partition that oversegments an image into convex cells, the algorithm refines the geometry of the partition while labeling its cells by a semantic class. The result is a set of polygons, each capturing an object in the image. The quality of a configuration is measured by an energy that accounts for both the fidelity to input data and the complexity of the output polygons. To efficiently explore the configuration space, we perform splitting and merging operations in tandem on the cells of the polygonal partition. The exploration mechanism is controlled by a priority queue that sorts the operations most likely to decrease the energy. We show the potential of our algorithm on different types of scenes, from organic shapes to man-made objects through floor maps, and demonstrate its efficiency compared to existing vectorization methods.

International audience; Converting point clouds generated by Laser scanning, multiview stereo imagery or depth cameras into compact polygon meshes is a challenging problem in vision. Existing methods are either robust to imperfect data or scalable, but rarely both. In this paper, we address this issue with an hybrid method that successively connects and slices planes detected from 3D data. The core idea consists in constructing an efficient and compact partitioning data structure. The later is i) spatially-adaptive in the sense that a plane slices a restricted number of relevant planes only, and ii) composed of components with different structural meaning resulting from a preliminary analysis of the plane connec-tivity. Our experiments on a variety of objects and sensors show the versatility of our approach as well as its competitiveness with respect to existing methods.

International audience; We present a method to create vector cliparts from photographs. Our approach aims at reproducing two key properties of cliparts: they should be easily editable, and they should represent image content in a clean, simplified way. We observe that vector artists satisfy both of these properties by modeling cliparts with linear color gradients, which have a small number of parameters and approximate well smooth color variations. In addition, skilled artists produce intricate yet editable artworks by stacking multiple gradients using opaque and semi-transparent layers. Motivated by these observations, our goal is to decompose a bitmap photograph into a stack of layers, each layer containing a vector path filled with a linear color gradient. We cast this problem as an optimization that jointly assigns each pixel to one or more layer and finds the gradient parameters of each layer that best reproduce the input. Since a trivial solution would consist in assigning each pixel to a different, opaque layer, we complement our objective with a simplicity term that favors decompositions made of few, semi-transparent layers. However, this formulation results in a complex combinatorial problem combining discrete unknowns (the pixel assignments) and continuous unknowns (the layer parameters). We propose a Monte Carlo Tree Search algorithm that efficiently explores this solution space by leveraging layering cues at image junctions. We demonstrate the effectiveness of our method by reverse-engineering existing cliparts and by creating original cliparts from studio photographs.

We introduce a novel approach that reconstructs 3D urban scenes in the form of levels of detail (LODs). Starting from raw datasets such as surface meshes generated by multiview stereo systems, our algorithm proceeds in three main steps: classification, abstraction, and reconstruction. From geometric attributes and a set of semantic rules combined with a Markov random field, we classify the scene into four meaningful classes. The abstraction step detects and regularizes planar structures on buildings, fits icons on trees, roofs, and facades, and performs filtering and simplification for LOD generation. The abstracted data are then provided as input to the reconstruction step which generates watertight buildings through a min-cut formulation on a set of 3D arrangements. Our experiments on complex buildings and large-scale urban scenes show that our approach generates meaningful LODs while being robust and scalable. By combining semantic segmentation and abstraction, it also outperforms general mesh approximation approaches at preserving urban structures.