检索结果-内蒙古大学图书馆

Deep reinforcement learning for motion planning of mobile robots

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Butyrev, Leonid Edelhäußer, Thorsten Mutschler, Christopher Fraunhofer Institute for Integrated Circuits Iis Precise Positioning and Analytics Department Machine Learning and Information Fusion Group Nuremberg Germany Computer Science Department Machine Learning and Data Analytics Lab Erlangen Germany

This paper presents a novel motion and trajectory planning algorithm for nonholonomic mobile robots that uses recent advances in deep reinforcement learning. Starting from a random initial state, i.e., position, velocity and orientation, the robot reaches an arbitrary target state while taking both kinematic and dynamic constraints into account. Our deep reinforcement learning agent not only processes a continuous state space it also executes continuous actions, i.e., the acceleration of wheels and the adaptation of the steering angle. We evaluate our motion and trajectory planning on a mobile robot with a differential drive in a simulation environment. Copyright © 2019, The Authors. All rights reserved.

关键词： Deep learning

learning with group Noise

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Wang, Qizhou Yao, Jiangchao Gong, Chen Liu, Tongliang Gong, Mingming Yang, Hongxia Han, Bo Department of Computer Science Hong Kong Baptist University Hong Kong Key Laboratory of Intelligent Perception and Systems for High-Dimensional Information of MoE School of Computer Science and Engineering Nanjing University of Science and Technology China Data Analytics and Intelligence Lab Alibaba Group China Department of Computing Hong Kong Polytechnic University Hong Kong Trustworthy Machine Learning Lab School of Computer Science University of Sydney Australia School of Mathematics and Statistics University of Melbourne Australia

machine learning in the context of noise is a challenging but practical setting to plenty of real-world applications. Most of the previous approaches in this area focus on the pairwise relation (casual or correlational relationship) with noise, such as learning with noisy labels. However, the group noise, which is parasitic on the coarse-grained accurate relation with the fine-grained uncertainty, is also universal and has not been well investigated. The challenge under this setting is how to discover true pairwise connections concealed by the group relation with its fine-grained noise. To overcome this issue, we propose a novel Max-Matching method for learning with group noise. Specifically, it utilizes a matching mechanism to evaluate the relation confidence of each object (cf. Figure 1) w.r.t. the target, meanwhile considering the Non-IID characteristics among objects in the group. Only the most confident object is considered to learn the model, so that the fine-grained noise is mostly dropped. The performance on a range of real-world datasets in the area of several learning paradigms demonstrates the effectiveness of Max-Matching. Copyright © 2021, The Authors. All rights reserved.

关键词： machine learning

Unmasking DeepFakes with simple features

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Durall, Ricard Keuper, Margret Pfreundt, Franz-Josef Keuper, Janis Fraunhofer ITWM Germany IWR University of Heidelberg Germany Fraunhofer Center Machine Learning Germany Data and Web Science Group University Mannheim Germany Institute for Machine Learning and Analytics Offenburg University Germany

Deep generative models have recently achieved impressive results for many real-world applications, successfully generating high-resolution and diverse samples from complex data sets. Due to this improvement, fake digital contents have proliferated growing concern and spreading distrust in image content, leading to an urgent need for automated ways to detect these AI-generated fake images. Despite the fact that many face editing algorithms seem to produce realistic human faces, upon closer examination, they do exhibit artifacts in certain domains which are often hidden to the naked eye. In this work, we present a simple way to detect such fake face images - so-called DeepFakes. Our method is based on a classical frequency domain analysis followed by a basic classifier. Compared to previous systems, which need to be fed with large amounts of labeled data, our approach showed very good results using only a few annotated training samples and even achieved good accuracies in fully unsupervised scenarios. For the evaluation on high resolution face images, we combined several public data sets of real and fake faces into a new benchmark: Faces-HQ. Given such high-resolution images, our approach reaches a perfect classification accuracy of 100% when it is trained on as little as 20 annotated samples. In a second experiment, in the evaluation of the medium-resolution images of the CelebA data set, our method achieves 100% accuracy supervised and 96% in an unsupervised setting. Finally, evaluating a low-resolution video sequences of the FaceForensics++ data set, our method achieves 91% accuracy detecting manipulated videos. Copyright © 2019, The Authors. All rights reserved.

关键词： Frequency domain analysis

Harnessing multimodal approaches for depression detection using large language models and facial expressions

学校读者我要写书评

暂无评论

Npj mental health research

Npj mental health research 2024年第1期3卷 66页

作者： Misha Sadeghi Robert Richer Bernhard Egger Lena Schindler-Gmelch Lydia Helene Rupp Farnaz Rahimi Matthias Berking Bjoern M Eskofier Machine Learning and Data Analytics Lab (MaD Lab) Department Artificial Intelligence in Biomedical Engineering (AIBE) Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) Erlangen 91052 Germany. misha.sadeghi@fau.de. Machine Learning and Data Analytics Lab (MaD Lab) Department Artificial Intelligence in Biomedical Engineering (AIBE) Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) Erlangen 91052 Germany. Chair of Visual Computing (LGDV) Department of Computer Science Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) Erlangen 91058 Germany. Chair of Clinical Psychology and Psychotherapy (KliPs) Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) Erlangen 91052 Germany. Translational Digital Health Group Institute of AI for Health Helmholtz Zentrum München - German Research Center for Environmental Health Neuherberg 85764 Germany.

Detecting depression is a critical component of mental health diagnosis, and accurate assessment is essential for effective treatment. This study introduces a novel, fully automated approach to predicting depression severity using the E-DAIC dataset. We employ Large Language Models (LLMs) to extract depression-related indicators from interview transcripts, utilizing the Patient Health Questionnaire-8 (PHQ-8) score to train the prediction model. Additionally, facial data extracted from video frames is integrated with textual data to create a multimodal model for depression severity prediction. We evaluate three approaches: text-based features, facial features, and a combination of both. Our findings show the best results are achieved by enhancing text data with speech quality assessment, with a mean absolute error of 2.85 and root mean square error of 4.02. This study underscores the potential of automated depression detection, showing text-only models as robust and effective while paving the way for multimodal analysis.

关键词：

ViPR: Visual-odometry-aided pose regression for 6DoF camera localization

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Ott, Felix Feigl, Tobias Löffler, Christoffer Mutschler, Christopher Nürnberg Germany Programming Systems Group FAU Erlangen-Nürnberg Germany Machine Learning and Data Analytics Lab FAU Erlangen-Nürnberg Germany

Visual Odometry (VO) accumulates a positional drift in long-term robot navigation tasks. Although Convolutional Neural Networks (CNNs) improve VO in various aspects, VO still suffers from moving obstacles, discontinuous observation of features, and poor textures or visual information. While recent approaches estimate a 6DoF pose either directly from (a series of) images or by merging depth maps with the optical flow (OF), research that combines absolute pose regression with OF is limited. We propose ViPR, a novel architecture for long-term 6DoF VO that leverages synergies between absolute pose estimates (from PoseNet-like architectures) and relative pose estimates (from FlowNet-based architectures) by combining both through recurrent layers. Experiments with known publicly available datasets and with our own Industry dataset show that our novel design outperforms existing techniques in long-term navigation tasks. Copyright © 2019, The Authors. All rights reserved.

关键词： Computer vision

Complementary App-Based Yoga Home Exercise Therapy for Patients With Axial Spondyloarthritis: Usability Study

学校读者我要写书评

暂无评论

JMIR Formative Research 2024年 8卷 e57185页

作者： Grube, Lara Petit, Pascal Vuillerme, Nicolas Nitschke, Marlies Nwosu, Obioma Bertrand Knitza, Johannes Krusche, Martin Seifer, Ann-Kristin Eskofier, Bjoern Schett, Georg Morf, Harriet Department of Internal Medicine 3- Rheumatology & Immunology Universitätsklinikum Erlangen Friedrich-Alexander-Universität Erlangen-Nürnberg Erlangen Germany Deutsches Zentrum Immuntherapie Universitätsklinikum Erlangen Friedrich-Alexander-Universität Erlangen-Nürnberg Erlangen Germany AGEIS Université Grenoble Alpes Grenoble France Institut Universitaire de France Paris France LabCom Telecom4Health Orange Labs & Université Grenoble Alpes CNRS Inria Grenoble France Machine Learning and Data Analytics Lab Department Artificial Intelligence in Biomedical Engineering (AIBE) Friedrich-Alexander-Universität Erlangen-Nürnberg Erlangen Germany Institute for Digital Medicine University Hospital of Giessen and Marburg Philipps-University Marburg Marburg Germany III. Department of Medicine University Medical Center Hamburg-Eppendorf Hamburg Germany Translational Digital Health Group Institute of AI for Health German Research Center for Environmental Health Helmholtz Zentrum München Neuherberg Germany

Background: Axial spondyloarthritis (AS) is a chronic inflammatory rheumatic disease characterized by potentially disabling inflammation of the spine and adjacent joints. Regular exercise is a cornerstone of treatment. However, patients with AS currently have little support. YogiTherapy (MaD Lab) is an app developed to support patients with AS by providing instructions for yoga-based home exercise therapy. Objective: This study aimed to evaluate the usability and acceptance of the newly designed YogiTherapy app for patients with AS. Methods: Patients completed the User Version of the Mobile Application Rating Scale (uMARS) and net promoter score (NPS) questionnaires after the app introduction. Wilcoxon Mann-Whitney rank sum test, chi-square test for count data, and correlation analysis were conducted to examine the usability of the app, acceptance, and patient characteristics. Results: A total of 65 patients with AS (33, 51% female;age: mean 43.3, SD 13.6 years) were included in the study from May 2022 to June 2023. Subsequently, the data were analyzed. Usability was rated moderate, with a mean uMARS of 3.35 (SD 0.47) points on a scale from 0 to 5. The highest-rated uMARS dimension was information (mean 3.88, SD 0.63), followed by functionality (mean 3.84, SD 0.87). Females reported a significantly higher uMARS total score than males (mean 3.47, SD 0.48 vs mean 3.23, SD 0.45;P=.03, Vargha and Delaney A [VDA] 0.66, 95% CI 0.53-0.77). The mean average of the NPS was 6.23 (SD 2.64) points (on a scale from 0 to 10), based on 43% (26/65 nonpromoters, 42% (25/65) indifferent, and 15% (9/65) promoters. A total of 7% (5/65) of those surveyed did not answer the question. When applying the NPS formula, the result is –26%. The NPS showed a positive correlation with the usage of mobile apps (r=0.39;P=.02). uMARS functionality was significantly higher rated by patients younger than 41 years (mean 4.17, SD 0.55 vs mean 3.54, SD 1;P<.001;VDA 0.69, 95% CI 0.56-0.80). Patients consi

关键词： adjacent joints ankylosing spondylitis app apps axial spondylarthritis chronic correlation analysis DHA digital health application digital therapeutics DTx eHealth exercise home exercise joint joints patient patient acceptance patients physical exercise rheumatic disease self-assessment spine spondyloarthritis therapy usability usability study Yoga YogiTherapy

A Digital Twin to overcome long-time challenges in Photovoltaics

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Lüer, Larry Peters, Marius Smith, Ana Sunčana Dorschky, Eva Eskofier, Bjoern M. Liers, Frauke Franke, Jörg Sjarov, Martin Brossog, Matthias Guldi, Dirk Maier, Andreas Brabec, Christoph J. Friedrich-Alexander-Universität Erlangen-Nürnberg Martensstrasse 7 Erlangen91058 Germany Immerwahrstraße 2 Erlangen91058 Germany PULS Group Department of Physics Friedrich-Alexander-Universität Erlangen-Nürnberg IZNF Erlangen Germany Division of Physical Chemistry Ruđer Bošković Institute Bijenička cesta 54 Zagreb Croatia Machine Learning and Data Analytics Lab Friedrich-Alexander-Universität Erlangen-Nürnberg Carl-Thiersch-Straße 2b Bayern Erlangen91052 Germany Friedrich-Alexander-Universität Erlangen-Nürnberg Cauerstr. 11 Erlangen91058 Germany Friedrich-Alexander-Universität Erlangen-Nürnberg Egerlandstr. 7 Erlangen91058 Germany Department of Chemistry and Pharmacy Friedrich-Alexander-Universität Erlangen-Nürnberg Egerlandstr. 3 Erlangen91058 Germany Informatics Department Friedrich-Alexander-Universität Erlangen-Nürnberg Martenstr. 3 Erlangen91058 Germany

The recent successes of emerging photovoltaics (PV) such as organic and perovskite solar cells are largely driven by innovations in material science. However, closing the gap to commercialization still requires significant innovation to match contradicting requirements such as performance, longevity and recyclability. The rate of innovation, as of today, is limited by a lack of design principles linking chemical motifs to functional microscopic structures, and by an incapacity to experimentally access microscopic structures from investigating macroscopic device properties. Both limitations in turn are caused by an individualist approach to learning, not being able to produce consistent datasets large enough to find patterns and physical laws leading us to breakthrough innovations. In this work, we envision a layout of a Digital Twin for PV materials aimed at removing both limitations. The layout combines machine learning approaches, as performed in materials acceleration platforms (MAPs), with mathematical models derived from the underlying physics and digital twin concepts from the engineering world. This layout will allow using high-throughput (HT) experimentation in MAPs to improve the parametrization of quantum chemical and solid-state models. In turn, the improved and generalized models can be used to obtain the crucial structural parameters from HT data. HT experimentation will thus yield a detailed understanding of generally valid structure-property relationships, enabling inverse molecular design, that is, predicting the optimal chemical structure and process conditions to build PV devices satisfying a multitude of requirements at the same time. After motivating our proposed layout of the digital twin with causal relationships in material science, we discuss the current state of the enabling technologies such as model-based inference of hidden parameters, already being able to yield insight from HT data today. However, we identify the multiscale nature of PV

关键词： machine learning

Epitopedia: identifying molecular mimicry between pathogens and known immune epitopes

学校读者我要写书评

暂无评论

ImmunoInformatics 2023年 9卷

作者： Christian A Balbin Janelle Nunez-Castilla Vitalii Stebliankin Prabin Baral Masrur Sobhan Trevor Cickovski Ananda Mohan Mondal Giri Narasimhan Prem Chapagain Kalai Mathee Jessica Siltberg-Liberles Department of Biological Sciences College of Arts Science and Education Florida International University Miami United States Bioinformatics Research Group (BioRG) Knight Foundation School of Computing and Information Sciences Florida International University Miami United States Department of Physics College of Arts Science and Education Florida International University Miami United States Machine Learning and Data Analytics Group (MLDAG) Knight Foundation School of Computing and Information Sciences Florida International University Miami United States Biomolecular Sciences Institute Florida International University Miami United States Department of Human and Molecular Genetics Herbert Wertheim College of Medicine Florida International University Miami United States

Upon infection, foreign antigenic proteins stimulate the host's immune system to produce antibodies targeting the pathogen. These antibodies bind to regions on the antigen called epitopes. Structural similarity (molecular mimicry) of epitopes between an infecting pathogen and host proteins or other pathogenic proteins the host has previously encountered can impact the host immune response to the pathogen and may lead to cross-reactive antibodies. The ability to identify potential regions of molecular mimicry in a pathogen can illuminate immune effects which are especially important to pathogen treatment and vaccine design. Here we present Epitopedia, a software pipeline that facilitates the identification of regions that may exhibit potential three-dimensional molecular mimicry between an antigenic pathogen protein and known immune epitopes as catalogued by the Immune Epitope database (IEDB). Epitopedia is open-source software released under the MIT license and is freely available on GitHub, including a Docker container with all other software dependencies preinstalled. We performed an analysis describing how various secondary structure states, identity between pentapeptide pairs, and identity between the parent sequences of pentapeptide pairs affects RMSD. We found that pentapeptides pairs in a helical conformation had considerably lower RMSD values than those in extended or coil conformations. We also found that RMSD is significantly increased when pentapeptide pairs are from non-homologous sequences.

关键词：

Analyzing the Structure of Attention in a Transformer Language Model

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Vig, Jesse Belinkov, Yonatan Palo Alto Research Center Machine Learning and Data Science Group Interaction and Analytics Lab Palo AltoCA United States Harvard John A. Paulson School of Engineering and Applied Sciences MIT Computer Science and Artificial Intelligence Laboratory CambridgeMA United States

The Transformer is a fully attention-based alternative to recurrent networks that has achieved state-of-the-art results across a range of NLP tasks. In this paper, we analyze the structure of attention in a Transformer language model, the GPT-2 small pretrained model. We visualize attention for individual instances and analyze the interaction between attention and syntax over a large corpus. We find that attention targets different parts of speech at different layer depths within the model, and that attention aligns with dependency relations most strongly in the middle layers. We also find that the deepest layers of the model capture the most distant relationships. Finally, we extract exemplar sentences that reveal highly specific patterns targeted by particular attention heads. Copyright © 2019, The Authors. All rights reserved.

关键词： Recurrent neural networks