A5102957644- Autonomous Vehicle Technology and Safety
- Video Surveillance and Tracking Methods
- Advanced Image and Video Retrieval Techniques
- Traffic Prediction and Management Techniques
- Advanced Neural Network Applications
- Computer Graphics and Visualization Techniques
- Advanced Vision and Imaging
- Reinforcement Learning in Robotics
- Human Pose and Action Recognition
- Food Supply Chain Traceability
- Generative Adversarial Networks and Image Synthesis
- Traffic control and management
- Image and Object Detection Techniques
- Image Processing and 3D Reconstruction
- Microbial infections and disease research
- Cerebral Palsy and Movement Disorders
- Animal Behavior and Welfare Studies
- Anomaly Detection Techniques and Applications
- Soft Robotics and Applications
- Industrial Vision Systems and Defect Detection
- Time Series Analysis and Forecasting
- Coagulation, Bradykinin, Polyphosphates, and Angioedema
- Oral and Maxillofacial Pathology
- Botulinum Toxin and Related Neurological Disorders
- Periodontal Regeneration and Treatments
University of Modena and Reggio Emilia
2017-2021
Ferrari (Italy)
2017-2019
Universidad de Las Palmas de Gran Canaria
2018
University of Ferrara
1996
Motivated by the impact of large-scale datasets on ML systems we present largest self-driving dataset for motion prediction to date, containing over 1,000 hours data. This was collected a fleet 20 autonomous vehicles along fixed route in Palo Alto, California, four-month period. It consists 170,000 scenes, where each scene is 25 seconds long and captures perception output system, which encodes precise positions motions nearby vehicles, cyclists, pedestrians time. On top this, contains...
In this work we present a simple end-to-end trainable machine learning system capable of realistically simulating driving experiences. This can be used for verification self-driving performance without relying on expensive and time-consuming road testing. particular, frame the simulation problem as Markov Process, leveraging deep neural networks to model both state distribution transition function. These are directly from existing raw observations need any handcrafting in form plant or...
Vehicle re-identification plays a major role in modern smart surveillance systems. Specifically, the task requires capability to predict identity of given vehicle, dataset known associations, collected from different views and cameras. Generally, it can be cast as ranking problem: probe image model needs rank all database images based on their similarities w.r.t image. In line with recent research, we devise metric learning that employs supervision local constraints. particular, leverage...
People re-identification task has seen enormous improvements in the latest years, mainly due to development of better image features extraction from deep Convolutional Neural Networks (CNN) and availability large datasets. However, little research been conducted on animal identification re-identification, even if this knowledge may be useful a rich variety different scenarios. Here, we tackle cattle exploiting CNN show how is poorly related with human one, presenting unique challenges that...
Visual observation of uncontrolled real-world behavior leads to noisy observations, complicated by occlusions, ambiguity, variable motion rates, detection and tracking errors, slow transitions between behaviors, etc.We show in this paper that reliable estimates long-term trends can be extracted given enough data, even though from individual frames may noisy.We validate concept using a new public dataset approximately 20+ million daytime pig observations over 6 weeks their main growth stage,...
T he overgrowth‐affected gingiva of patients treated with cyclosporin A after kidney transplant was examined ultrastructural and histochemical methods to evaluate the involvement connective tissue. Gingival overgrowth has same clinical signs as local edema. The study showed that dimensional increase largely due increased production amorphous ground substance by fibroblasts, possibly resulting from an release histamine mast cells. data revealed affected tissues contained higher levels...
Diplegia is a specific subcategory of the wide spectrum motion disorders gathered under name cerebral palsy. Recent works proposed to use gait analysis for diplegia classification paving way automated analysis. A clinically established gait-based system divides diplegic patients into 4 main forms, each one associated with peculiar walking pattern. In this work, we apply two different deep learning techniques, namely, multilayer perceptron and recurrent neural networks , automatically...
Abstract Diseases of the respiratory system are known to negatively impact profitability pig industry, worldwide. Considering relatively short lifespan pigs, lesions can be still evident at slaughter, where they usefully recorded and scored. Therefore, slaughterhouse represents a key check-point assess health status providing unique valuable feedback farm, as well an important source data for epidemiological studies. Although relevant, scoring in slaughtered pigs very time-consuming costly...
Despite promising progress in reinforcement learning (RL), developing algorithms for autonomous driving (AD) remains challenging: one of the critical issues being absence an open-source platform capable training and effectively validating RL policies on real-world data. We propose DriverGym, OpenAI Gym-compatible environment specifically tailored driving. DriverGym provides access to more than 1000 hours expert logged data also supports reactive data-driven agent behavior. The performance...
In this article we introduce a new self-supervised, semi-parametric approach for synthesizing novel views of vehicle starting from single monocular image. Differently <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">parametric</i> (i.e., entirely learning-based) methods, show how xmlns:xlink="http://www.w3.org/1999/xlink">a-priori</i> geometric knowledge about the object and 3D world can be successfully integrated into deep learning based...
Paroxysmal nocturnal hemoglobinuria (PNH) is a clonal nonneoplastic hematopoietic stem cell disease characterized by an acquired mutation of the PIG-A gene with reduction or absence CD55 and CD59. The these proteins renders PNH erythrocytes susceptible to complement-mediated hemolysis. We report case patient before during pregnancy until delivery. observed treated some postpartum thrombotic complications. Eculizumab should be used caution in pregnancy. There are several reports supporting...
In this work we propose a deep learning pipeline to predict the visual future appearance of an urban scene. Despite recent advances, generating entire scene in end-to-end fashion is still far from being achieved. Instead, here follow two stages approach, where interpretable information included loop and each actor modelled independently. We leverage per-object novel view synthesis paradigm; i.e. synthetic representation object undergoing geometrical roto-translation 3D space. Our model can...
To achieve robustness in Re-Identification, standard methods leverage tracking information a Video-To-Video fashion. However, these solutions face large drop performance for single image queries (e.g., Image-To-Video setting). Recent works address this severe degradation by transferring temporal from Video-based network to an Image-based one. In work, we devise training strategy that allows the transfer of superior knowledge, arising set views depicting target object. Our proposal - Views...