Refine
Year of publication
- 2020 (96) (remove)
Document Type
- Article (47)
- Conference Proceeding (35)
- Book (4)
- Part of a Book (4)
- Report (2)
- Working Paper (2)
- Doctoral Thesis (1)
- Other (1)
Language
- English (96) (remove)
Keywords
- GIS (2)
- Values Management (2)
- remote sensing (2)
- windthrow (2)
- Aggregation, Biomolecular self-assembly, Electrostatic interactions, Charged colloids, Colloids, Proteins, Density functional calculations, Polymers & Soft Matter, Biological Physics, Statistical Physics (1)
- Allgemeines Gleichberechtigungsgesetz (1)
- CNNs (1)
- CSR (1)
- Christian employers (1)
- Corporate Ethics (1)
Institute
- Fakultät Informatik und Wirtschaftsinformatik (19)
- Fakultät Angewandte Sozialwissenschaften (10)
- Fakultät Elektrotechnik (8)
- Fakultät für angewandte Natur- und Geisteswissenschaften (5)
- Fakultät Wirtschaftsingenieurwesen (4)
- Fakultät Wirtschaftswissenschaften (4)
- Institut Digital Engineering (IDEE) (4)
- Technologie TransferZentrum-Elektromobilität (TTZ-EMO) (4)
- Fakultät Kunststofftechnik und Vermessung (3)
- Fakultät Maschinenbau (1)
Acknowledging that the climate emergency is a transdisciplinary challenge that changes everything, this paper offers reflections on teaching and research, attending to the public and students' demands, witnessed at a global climate protest- ‘to not remain as bystanders, but join in action!‘ It asks, what could we begin to do to teach, research and write for climate action, and for solidarity? We offer a joint reflection on the experience of a collaborative Erasmus+ Disciplinary Excellence in Teaching, Learning and Assessment (Erasmus DELTA) training exchange and consider the challenges and benefits for engaged scholarship, subject specific curriculum enhancement and creative pedagogy.
Petition 0429/2017 has been brought forward by Prof. Harald Bolsinger. In this article he describes backgrounds and developments around the petition. The petition, submitted on May the 8th 2017, addressed the investments policy of the ECB. Compliance with EU Fundamental Rights, it claimed, should be specifically included into the eligibility for EU owned assets. As Bolsinger explains, at the moment the ECB is involved in violations of the Charter of Fundamental Rights of the European Union through possession and trading of unethical securities.
Camera Re-Localization with Data Augmentation by Image Rendering and Image-to-Image Translation
(2020)
Self-localization of cars, robots or Unmanned Aerial Vehicles (UAVs) as well as self-localization of pedestrians is and will be of high interest for a wide range of applications.
A major task is autonomous navigation of vehicles, whereas the localization in the surrounding scene is a key component.
Since cameras are well-established built-in sensors in cars, robots and UAVs, there is little to no extra cost in utilizing them for subsequent localization.
The same applies for pedestrian localization, where smartphones serve as mobile platforms for cameras.
Camera re-localization, where the pose of a camera is determined with respect to a certain scene, is therefore a valuable process to solve or support localization solutions of such vehicles or pedestrians.
Cameras are low-cost sensors that are established commonly in the everyday life of humans and machines.
The support of camera re-localization is not limited to applications related to navigation but can in general be applied to support image analysis or image processing tasks as scene reconstruction, detection, classification or alike.
For these purposes, this thesis concerns the improvement of the camera re-localization task.
As Convolutional Neural Networks (CNNs) and hybrid pipelines to regress camera poses are recently competing against established hand-crafted designed pipelines reaching similar or superior accuracies, the focus is set on the former in this thesis.
The main contributions of this thesis include the design of a CNN to regress camera poses, with focus on a lightweight architecture fitting the requirements to be applicable on mobile platforms.
This network achieves accuracies in the same order as CNNs with larger model sizes.
Furthermore, the performance of CNNs is highly depending on the quantity and quality of training data utilized for their optimization.
Hence, further contributions are considering image rendering and image-to-image translation to extend such training data in terms of Data Augmentation (DA).
3D models are utilized for image rendering to valuable extend training datasets.
Generative Adversarial Networks (GANs) serve for DA by image-to-image translation. Whereas image rendering is increasing the quantity of images in datasets, image-to-image translation on the other hand aims to enhance the quality of such data.
Experiments are carried out on datasets augmented by image rendering and image-to-image translation. It is shown, that both approaches valuable enhance the localization concerning accuracy.
Therefore, state-of-the-art localization is improved by DA in this thesis
One of the major shortcomings of variational autoencoders is the inability to produce generations from the individual modalities of data originating from mixture distributions. This is primarily due to the use of a simple isotropic Gaussian as the prior for the latent code in the ancestral sampling procedure for the data generations. We propose a novel formulation of variational autoencoders, conditional prior VAE (CP-VAE), which learns to differentiate between the individual mixture components and therefore allows for generations from the distributional data clusters. We assume a two-level generative process with a continuous (Gaussian) latent variable sampled conditionally on a discrete (categorical) latent component. The new variational objective naturally couples the learning of the posterior and prior conditionals, and the learning of the latent categories encoding the multimodality of the original data in an unsupervised manner. The data-dependent conditional priors are then used to sample the continuous latent code when generating new samples from the individual mixture components corresponding to the multimodal structure of the original data. Our experimental results illustrate the generative performance of our new model comparing to multiple baselines.
One of the major shortcomings of variational autoencoders is the inability to produce generations from the individual modalities of data originating from mixture distributions. This is primarily due to the use of a simple isotropic Gaussian as the prior for the latent code in the ancestral sampling procedure for data generations. In this paper, we propose a novel formulation of variational autoencoders, conditional prior VAE (CP-VAE), with a two-level generative process for the observed data where continuous 𝐳 and a discrete 𝐜 variables are introduced in addition to the observed variables 𝐱. By learning data-dependent conditional priors, the new variational objective naturally encourages a better match between the posterior and prior conditionals, and the learning of the latent categories encoding the major source of variation of the original data in an unsupervised manner. Through sampling continuous latent code from the data-dependent conditional priors, we are able to generate new samples from the individual mixture components corresponding, to the multimodal structure over the original data. Moreover, we unify and analyse our objective under different independence assumptions for the joint distribution of the continuous and discrete latent variables. We provide an empirical evaluation on one synthetic dataset and three image datasets, FashionMNIST, MNIST, and Omniglot, illustrating the generative performance of our new model comparing to multiple baselines.
Lifelong generative modeling
(2020)
Lifelong learning is the problem of learning multiple consecutive tasks in a sequential manner, where knowledge gained from previous tasks is retained and used to aid future learning over the lifetime of the learner. It is essential towards the development of intelligent machines that can adapt to their surroundings. In this work we focus on a lifelong learning approach to unsupervised generative modeling, where we continuously incorporate newly observed distributions into a learned model. We do so through a student-teacher Variational Autoencoder architecture which allows us to learn and preserve all the distributions seen so far, without the need to retain the past data nor the past models. Through the introduction of a novel cross-model regularizer, inspired by a Bayesian update rule, the student model leverages the information learned by the teacher, which acts as a probabilistic knowledge store. The regularizer reduces the effect of catastrophic interference that appears when we learn over sequences of distributions. We validate our model’s performance on sequential variants of MNIST, FashionMNIST, PermutedMNIST, SVHN and Celeb-A and demonstrate that our model mitigates the effects of catastrophic interference faced by neural networks in sequential learning scenarios.
Protocol for the management and the monitoring of the measurements for the COVID19 containment
(2020)
Objective:
Behavioral activation constitutes a promising behavioral treatment for depression. Due to its contextual and idiographic approach, the intervention might be well suited for treating depression in culturally diverse populations.
Method:
The authors conducted a systematic literature review on culturally adapted behavioral activation treatments.
Results:
Seventeen studies were identified through database searching involving different target populations and a variety of adapted interventions. Circumstances were frequently shaped by cultural values and a wide range of environmental stressors. Adaptations were found in different dimensions including language, content, methods, and context. Across studies, results indicated the effectiveness of behavioral activation and its cultural adaptations for treating depression in their respective target groups.
Discussion:
The results of this review may serve as an input both for practitioners employing behavioral activation in their daily work with culturally diverse clients and for researchers interested in culturally adapting treatment to specific populations.
XDose: toward online cross-validation of experimental and computational X-ray dose estimation
(2020)
There are novel structures of society such as Industry 4.0 and Connected Industries. Under the structures, a novel safety management system considering man-machine interaction is very important topic. Safety 2.0, a new safety control, is proposed from Japan. It seems that direct and quantitative validity of risk assessment and risk reduction measure such as reinforcement of safety behavior and/or reduction of unsafe behavior is now not sufficiently accomplished. Some quantitative assessment for residual risks focusing on worker's behavior is necessary to make Vision Zero function better. We introduce to a principle of Behavior-Based Safety (BBS) which might be applicable to reduce residual risk for safety of workers. BBS, focused on behavioral modification using principle of reinforcement, is one group of Behavior Analysis of Psychology. Goals of BBS is prediction, control and quantitative measurement of target behavior.
Purpose:
Many students have difficulties in public speaking because of their use of filled pauses (e.g., utterances like “um”; misuse of the word “like”). Mancuso and Miltenberger used habit reversal to decrease filled pauses in public speaking. The present study aimed at replicating this study as a student project.
Method:
Participants were four undergraduate students of social work. The training phase and total number of sessions were shortened compared to the original study.
Results:
The mean number of responses (filled pauses) per minute decreased throughout the study and during follow-up measurement.
Discussion:
Nevertheless, the replication of experimental control wasn’t successful because three of the four participants showed a decrease in response frequency already during the baseline. As a project, the study demonstrates that students of social work with only rudimentary training in single systems research methods can implement evidence-based practice procedures in their work with clients.
Forest damage due to storms causes economic loss and requires a fast response to prevent further damage such as bark beetle infestations. By using Convolutional Neural Networks (CNNs) in conjunction with a GIS, we aim at completely streamlining the detection and mapping process for forest agencies. We developed and tested different CNNs for rapid windthrow detection based on PlanetScope satellite data and high-resolution aerial image data. Depending on the meteorological situation after the storm, PlanetScope data might be rapidly available due to its high temporal resolution, while the acquisition of high-resolution airborne data often takes weeks to a month and is, therefore, used in a second step for more detailed mapping. The study area is located in Bavaria, Germany (ca. 165 km2), and labels for damaged areas were provided by the Bavarian State Institute of Forestry (LWF). Modifications of a U-Net architecture were compared to other approaches using transfer learning (e.g., VGG19) to find the most efficient architecture for the task on both datasets while keeping the computational time low. A custom implementation of U-Net proved to be more accurate than transfer learning, especially on medium (3 m) resolution PlanetScope imagery (intersection over union score (IoU) 0.55) where transfer learning completely failed. Results for transfer learning based on VGG19 on high-resolution aerial image data are comparable to results from the custom U-Net architecture (IoU 0.76 vs. 0.73). When using both architectures on a dataset from a different area (located in Hesse, Germany), however, we find that the custom implementations have problems generalizing on aerial image data while VGG19 still detects most damage in these images. For PlanetScope data, VGG19 again fails while U-Net achieves reasonable mappings. Results highlight the potential of Deep Learning algorithms to detect damaged areas with an IoU of 0.73 on airborne data and 0.55 on Planet Dove data. The proposed workflow with complete integration into ArcGIS is well-suited for rapid first assessments after a storm event that allows for better planning of the flight campaign followed by detailed mapping in a second stage.
Malaria is one of the most cited vector-borne infectious diseases by climate change expert panels. Malaria vectors often need water sheets or wetlands to complete the disease life cycle. The current context of population mobility and global change requires detailed monitoring and surveillance of malaria in all countries. This study analysed the spatiotemporal distribution of death and illness cases caused by autochthonous and imported malaria in Spain during the 20th and 21st centuries using multidisciplinary sources, Geographic Information System (GIS) and geovisualisation. The results obtained reveal that, in the 20th and 21st centuries, malaria has not had a homogeneous spatial distribution. Between 1916 and 1930, 77% of deaths from autochthonous malaria were concentrated in only 20% of Spanish provinces; in 1932, 88% of patients treated in anti-malarial dispensaries were concentrated in these same provinces. These last data reveal the huge potential that anti-malarial dispensaries could have as a tool to reconstruct historical epidemiology. Spanish autochthonous malaria has presented epidemic upsurge episodes, especially those of 1917–1922 and 1939–1944, influenced by armed conflict, population movement and damaged health and hygiene conditions. Although meteorological variables have not played a key role in these epidemic episodes, they contributed by providing suitable conditions for their intensification. After the eradication of autochthonous malaria in 1961, imported malaria cases began to be detected in 1973, reaching more than 700 cases per year at the end of the second decade of the 21st century. Therefore, consistent and detailed historical studies are necessary to better understand the drivers that have led to the decline and elimination of malaria in Europe and other temperate countries.