We present City Of Light, a Unity-based, city-scale 116 km² simulator of Paris for high-throughput embodied AI research. COL fuses open geographic information system sources into geo-anchored, per-tile meshes and provides a configurable, stochastic runtime with controllable traffic and pedestrians. Agents receive frame-synchronized multi-sensor observations (RGB, depth, normals, semantics) and execute step-synchronized actions to navigate the environment. To support high-rate vision pipelines, we introduce TURBO, a Unity-Python bridge that streams multi-camera observations and allows control at up to 1300 FPS, achieving higher throughput than ML-Agents in our benchmark. We also provide a Street View Digital Twin that aligns simulator viewpoints with corresponding real-world panoramas for frame-accurate visual comparison and quantitative matching. COL enables fast scripting, large-scale data collection, and reinforcement learning in geo-anchored urban settings.

Chemical staining methods, while reliable, are time consuming and can be resource-intensive, involving costly chemical reagents and raising environmental concerns. This underscores the compelling need for alternative solutions such as virtual staining, which not only accelerates the diagnostic process but also enhances the flexibility of stain applications without the associated physical and chemical costs. Generative artificial intelligence technologies prove to be immensely useful in addressing these challenges. However, in healthcare, particularly within computational pathology, the high-stakes nature of decisions complicates the adoption of these tools due to their often opaque processes. Our work introduces an innovative approach that harnesses generative models for virtual stain transformations, improving performance, trustworthiness, scalability, and adaptability within computational pathology. The core of the proposed methodology involves a singular Hematoxylin and Eosin (H&E) encoder that supports multiple stain decoders. This design prioritizes critical regions in the latent space of H&E tissues, leading to a richer representation that enables precise synthetic stain generation by the decoders. Tested to simultaneously generate eight different stains from a single H&E slide, our method also offers significant scalability benefits for routine use by loading only necessary model components during production. We integrate label-free knowledge during training, using loss functions and regularization to minimize artifacts, thereby enhancing the accuracy of virtual staining in both paired and unpaired settings. To build trust in these synthetic stains, we employ a real-time self-inspection methodology using trained discriminators for each stain type, providing pathologists with confidence heatmaps to aid in their evaluations. In addition, we perform automatic quality checks on new H&E slides to ensure that they conform to the trained H&E distribution, guaranteeing the generation of high-quality synthetic stained slides. Recognizing the challenges pathologists face in adopting new technologies, we have encapsulated our method in an open-source, cloud-based proof-of-concept system. This system enables users to easily and virtually stain their H&E slides through a browser, eliminating the need for specialized technical knowledge and addressing common hardware and software challenges. It also facilitates real-time user feedback integration. Lastly, we have curated a novel dataset comprising eight different paired H&E/stains related to pediatric Crohn's disease at diagnosis, providing 30 whole slide images (WSIs) for each stain set (total of 480 WSIs) to stimulate further research in computational pathology.

Corn is cultivated by smallholder farmers in Ancash - Peru and it is one of the most important crops of the region. Climate change and migration from rural to urban areas are affecting agricultural production and therefore, food security. Information about the cultivated extension is needed for the authorities in order to evaluate the impact in the region. The present study proposes corn areas segmentation in multi-spectral images acquired from Unmanned Aerial Vehicles (UAV), using convolutional neural networks. U-net and U-net using VGG11 encoder were compared using dice and IoU coefficient as metrics. Results show that with the second model, 81.5% dice coefficient can be obtained in this challenging task, allowing envisaging an effective and efficient use of this technology, in this hard context.

In this paper we present a pipeline for automatic analysis of neuronal morphology: from detection, modeling to digital reconstruction. First, we present an automatic, unsupervised object detection framework using stochastic marked point process. It extracts connected neuronal networks by fitting special configuration of marked objects to the centreline of the neurite branches in the image volume giving us position, local width and orientation information. Semantic modeling of neuronal morphology in terms of critical nodes like bifurcations and terminals, generates various geometric and morphology descriptors such as branching index, branching angles, total neurite length, internodal lengths for statistical inference on characteristic neuronal features. From the detected branches we reconstruct neuronal tree morphology using robust and efficient numerical fast marching methods. We capture a mathematical model abstracting out the relevant position, shape and connectivity information about neuronal branches from the microscopy data into connected minimum spanning trees. Such digital reconstruction is represented in standard SWC format, prevalent for archiving, sharing, and further analysis in the neuroimaging community. Our proposed pipeline outperforms state of the art methods in tracing accuracy and minimizes the subjective variability in reconstruction, inherent to semi-automatic methods.

Automated algorithms to build accurate models of 3D neuronal arborization is much in demand due to large volume of microscopy data. We present a tracking algorithm for automatic and reliable extraction of neuronal morphology. It is robust to ambiguous branch discontinuities, variability of intensity and curvature of fibres, arbitrary branch cross-sections, noise and irregular background illumination. We complete the presentation of our method with demonstration of its performance on synthetic data modeling challenging scenarios and on confocal microscopy data of Olfactory Projection fibres from DIADEM data set with promising results.

In this paper we interest ourselves to the problem of flow tracking for dense crowds. For this purpose, we use a cloud of particles spread on the image according to the estimated crowd density and driven by the optical flow. This cloud of particles is considered as statistically representative of the crowd. Therefore, each particle has physical properties that enable us to assess the validity of its behavior according to the one expected from a pedestrians and to optimize its motion dictated by the optical flow. This leads us to three applications described in this paper: the detection of the entry and exit areas of the crowd in the image, the detection of dynamic occlusions and the possibility to link entry areas with exit ones according to the flow of the pedestrians. We provide the results of our experimentation on synthetic data and show promising results.

The extraction of nuclei from Haematoxylin and Eosin (H&E) stained biopsies present a particularly steep challenge in part due to the irregularity of the high-grade (most malignant) tumors. To your best knowledge, although some existing solutions perform adequately with relatively predictable low-grade cancers, solutions for the problematic high-grade cancers have yet to be proposed. In this paper, we propose a method for the extraction of cell nuclei from H&E stained biopsies robust enough to deal with the full range of histological grades observed in daily clinical practice. The robustness is achieved by combining a wide range of information including color, texture, scale and geometry in a multi-stage, Support Vector Machine (SVM) based framework to replace the original image with a new, probabilistic image modality with stable characteristics. The actual extraction of the nuclei is performed from the new image using Mark Point Processes (MPP), a state-of-the-art stochastic method. An empirical evaluation on clinical data provided and annotated by pathologists shows that our method greatly improves detection and extraction results, and provides a reliable solution with high grade cancers. Moreover, our method based on machine learning can easily adapt to specific clinical conditions. In many respects, our method contributes to bridging the gap between the computer vision technologies and their actual clinical use for breast cancer grading.

The incorporation of prior-knowledge into support vector machines (SVM) in order to compensate for inadequate training data has been the focus of previous research works and many found a kernel-based approach to be the most appropriate. However, they are more adapted to deal with broad domain knowledge (e.g. ``sets are invariant to permutations of the elements'') rather than task-specific properties (e.g. ``the weight of a person is cubically related to her height''). In this paper, we present the partially RBF (pRBF) kernels, our original framework for the incorporation of prior-knowledge about correlation patterns between specific features and the output label. pRBF kernels are based upon the tensor-product combination of the standard radial basis function (RBF) kernel with more specialized kernels and provide a natural way for the incorporation of a commonly available type of prior-knowledge. In addition to a theoretical validation of our framework, we propose an detailed empirical evaluation on real-life biological data which illustrates its ease-of-use and effectiveness. Not only pRBF kernels were able to improve the learning results in general but they also proved to perform particularly well when the training data set was very small or strongly biased, significantly broadening the field of application of SVMs.

The study of stem cells is one of the current most important biomedical research field. Understanding their development could allow multiple applications in regenerative medicine. For this purpose, we need automated methods for the segmentation and the modeling of neural stem cell development process into a neurosphere colony from phase contrast microscopy. We use such methods to extract relevant structural and textural features like cell division dynamism and cell behavior patterns for biological interpretation. The combination of phase contrast imaging, high fragility and complex evolution of neural stem cells pose many challenges in image processing and image analysis. This study introduces an on-line analysis method for the modeling of neurosphere evolution during the first three days of their development. From the corresponding time-lapse sequences, we extract information from the neurosphere using a combination of fast level set and curve detection for segmenting the cells. Then, based on prior biological knowledge, we generate possible and optimal 3-dimensional configuration using registration and evolutionary optimisation algorithm.

SVMs with the general purpose RBF kernel are widely considered as state-of-the-art supervised learning algorithms due to their effectiveness and versatility. However, in practice, SVMs often require more training data than readily available. Prior-knowledge may be available to compensate this shortcoming provided such knowledge can be effectively passed on to SVMs. In this paper, we propose a method for the incorporation of prior-knowledge via an adaptation of the standard RBF kernel. Our practical and computationally simple approach allows prior-knowledge in a variety of forms ranging from regions of the input space as crisp or fuzzy sets to pseudo-periodicity. We show that this method is effective and that the amount of required training data can be largely decreased, opening the way for new usages of SVMs. We propose a validation of our approach for pattern recognition and classification tasks with publicly available datasets in different application domains.

A consciousness-driven visual attention model is presented in this paper. It is based on a bio-inspired computer fovea model and a hierarchical analysis for the given visual receptive field. Indeed, the bio-inspired computer fovea model is used to simulate the neural activity on the human visual system, and the hierarchical analysis provides a function to explore the information on the given visual scene. The proposed model can evaluate the contents on the scene and automatically highlights visually important regions. This model can be used in various applications, such as surveillance, visual navigation and target acquisition, etc.

Most research works focus on pattern recognition within a small sample images but strategies for running efficiently these algorithms over large images are rarely if ever specifically considered. In particular, the new generation of satellite and microscopic images are acquired at a very high resolution and a very high daily rate. We propose an efficient, generic strategy to explore large images by combining computational geometry tools with a local signal measure of relevance in a dynamic sampling framework. An application to breast cancer grading from huge histopathological images illustrates the benefit of such a general strategy for new major applications in the field of microscopy.

Recent work in the field of machine translation (MT) evaluation suggests that sentence level evaluation based on machine learning (ML) can outperform the standard metrics such as BLEU, ROUGE and METEOR. We conducted a comprehensive empirical study on support vector methods for ML-based MT evaluation involving multi-class support vector machines (SVM) and support vector regression (SVR) with different kernel functions. We empathize on a systematic comparison study of multiple feature models obtained with feature selection and feature extraction techniques. Besides finding the conditions yielding the best empirical results, our study supports several unobvious conclusions regarding qualitative and quantitative aspects of feature sets in MT evaluation.

Bio-inspired computer vision is an emerging field. It aims to reproduce the capabilities of biological vision systems, eventually to simulate the visual functions for various purposes. In this paper, we propose a bio-inspired computer visual system using Graphical Processing Unit (GPU), and its application on breast cancer prognosis. The system extracts visual features from an input image using a mechanism which is similar to human visual system. Then classify the visual features using a machine learning algorithm. As a result, the elements of the input image, which might be related to particular knowledge concepts, can be identified.

In this paper, A configuration of Cellular Neural Network (CNN) is introduced to implement Principal Component Analysis (PCA). CNN is a parallel computing paradigm. Many researchers considered it as the next generation universal machine and developed so-called CNN universal chips. Based on the capability of CNN, an alternative PCA implementation named Principal Component Analyzing Cellular Neural Network (PCACNN) is proposed. PCA is used to reduce the dimensions of a given dataset in order to extract the principal information of the given dataset. In decades, many researchers presented their investigations based on PCA in order to improve the performance and/or to attack some open issues in specific fields. In this paper, PCA is implemented based on the architecture and capabilities of CNN. Consequently, the computing performance of PCA can be improved as long as the CNN architecture can be realized.

Automatic classification of working areas in peripheral blood smears can provide objective and reproducible quality control for the evaluation of smears and smear maker devices. However, it has drawn little research attention. In this paper we study this topic using image analysis and statistical pattern recognition methods. We employ generic features without requiring the extraction of individual cells. Two new spatial distribution features across scales are defined and utilized to classify working areas. We demonstrate that the only feature and method proposed in a similar work by others is insufficient to characterize the goodness of working areas, particularly the cell distribution. However, by utilizing it together with the features developed in this paper, we can achieve much better results. Our method has been tested on about 150 labeled images acquired from three malaria-infected Giemsa-stained blood smears using an oil immersion 100x objective lens.

Image segmentation has been studied for many years. But what factors influence segmentation results indeed? Why some images are easy to be handled while the others are not? In this paper we put forward the so-called 'image structure constant' and 'image structure map' to judge the complexity of an image. They can be applied on any image. 'Structure constant' can be found by a hierarchal segmentation method based on k-means and gray histogram, which is processed by increasing the clustering centers' number of k-means step by step and tracing the regions' change. At the same time its structure map can be formed reflecting the relationship between pixel gray values and image regions. With the structure constant and structure map we can dissert an image is easy to be segmented or not, quantitatively. Furthermore, a neighbor-matched-region (NMR) graph is designed to judge an image's complexity. Experiments show that the proposed concepts and the relevant algorithms are useful tools in analyzing images.

In this paper, we propose a unitary tool for modeling and analysis of discrete event systems monitoring. Uncertain knowledge of such tasks asks specific reasoning and adapted fuzzy logic modeling and analysis methods. In this context, we propose a new fuzzy Petri net called Fuzzy Reasoning Petri Net: the FRPN. The modeling consists in a set of two collaborative FRPN. The first is used for the fault dynamic state of the system by temporal spectrum of the marking. A monitoring fuzzy Petri net (MFPN) represents the fault tree. The second model, the recovery fuzzy Petri net (RFPN) corresponds to recovering activities. The two proposed models form a dynamic loop for production system monitoring and recovery. Production system is supposed to be modelized using temporal Petri nets, with the assumption that the primary fault symptoms are detected. These symptoms are considered in the MFPN and evolve according to all other derived faults of the system. Synchronizing signals, corresponding to different warning levels and alarms, form the interface between these two tools.

This paper introduces a modular modeling approach for distributed production systems, considering the production and maintenance processes synchronization. The production job shop, preventive and curative specialized tele-maintenance actions are studied. Stochastic synchronized Petri nets are used for modeling. This tool was adapted to the tele-maintenance distributed case and to integrate distant communications, synchronization problems and execution scheduling. Performance evaluation was performed using Markov Processes and Monte Carlo simulation.

This paper introduces the Recurrent Radial Basis Function networks (RRBF) for recognition of simple temporal sequence. The RRBF combines features from recurrent neural network and Radial Basis Function networks (RBF). An application has been developed by implementation on the IBM/ZISC (Zero Instruction Set Computer) card for temporal sequences recognition.

Stochastic Petri nets are a powerful tool for performance evaluation of concurrent systems like parallel computing, communication network and production systems. In many practical applications, performance evaluation using this model is very difficult because of the great dimension of the marking space. We present a graphical method for the reduction of stochastic Petri nets, applied to safe production system modeling. The approach is based on the principle of places interactivity in the model and the function of transition firing rates. The reduction of the Petri net is applied directly to the graphical model after a simple analysis of places efficiency by using mathematical techniques of birth-death processes. Thus, the problem of the model dimension is solved since our method is independent of the marking graph.

Our work concerns the adaptation of the singular perturbation method jointly to the homographic transformation to the category of ergodic Markov chains which presents the two-time-scale property. For Markov chains, the two-time-scale property becomes a property of two-weighting-scale of the states in the system evolution. This lead us to call the slow and fast parts of a decomposed system strong and respectively weak. The limit resolution methodology of the Markov chains by the method of singular perturbation assumes firstly the detection of the irreducible classes of the chain, and secondly, the decomposition of each final ergodic classes presenting the two-weighting-scale property. In the resolution at the limit of the decomposed system, we struck the problem of the stochasticity of the subsystems obtained using directly the singular perturbation method. Indeed, the strong and weak submatrix are not stochastic matrix. In our method, we use the homographic transformation in order to make stochastic the strong part matrix.>