The Augmented Vision Department of DFKI led by Prof. Dr. Didier Stricker, offers a student assistant job (part-time) for curious and passionate students, who want to develop themselves in Advanced Computer Vision.

Your Task

  • Researching and developing techniques for hand pose estimation.
  • Researching and developing techniques for hand-object interaction.
  • Implementing state-of-the-art methods to solve real-world problems such as gesture recognition and hand mesh reconstruction for AR/VR use-cases.

Your Qualifications

  • Good knowledge of Python and Pytorch
  • Interest in Deep Learning and Computer Vision
  • Master or high semester Bachelor

Your Benefits

  • Acquire skills in the domains of hand pose estimation, hand mesh reconstruction and hand-object interaction.
  • Opportunity to produce novel research work in the domain of hand-object interaction.
  • Opportunity to start your thesis with us.
  • Practical experience in modern Deep Learning techniques

Apply latest by: 15.06.2024

Please feel free to contact us if you have any questions regarding this position:

Christen.Millerdurai@dfki.de

Room 1.21, DFKI-Kaiserslautern

Arbeitgeber: RPTU

Befristung: Zunächst 2 Jahre, Verlängerung möglich

Beginn: 01.05.2024

Standort: Kaiserslautern

Umfang: Teilzeit, 75 %

Vergütung: E13, TV-L

Fachbereich: Informatik

Die Arbeitsgruppe Augmented Vision wird von Prof. Dr. Didier Stricker geleitet und ist Teil der Fachbereichs Informatik der RPTU und Teil des Deutschen Forschungszentrums für Künstliche Intelligenz (DFKI). In laufenden Forschungsprojekten befassen wir uns unter anderem mit nachhaltigen, energieeffizienten, und adaptiven Lernverfahren unter Verwendung von Event-Kameras und Spiking Neural Networks (SNNs). Zum nächst möglichen Zeitpunkt suchen wir Verstärkung für unser Team um Forschungsfragen rund um die dynamische Adaption von neuromorphen Sensoren unter Verwendung von neuronalen Netzen zu beantworten.

Ihr Aufgabengebiet

  • Grundlagen- und anwendungsorientiere Forschung im Bereich Neuromorphic Computing und Sensing
  • Sensor-Simulation in vorhanden Frameworks; Implementierung von adaptiven Sensoren
  • Arbeit in einem dynamischen, internationalen, und innovativen Arbeitsumfeld mit exzellenter Ausstattung
  • Simulation von Event-Daten und Training von Neuronalen Netzen mit diesen Daten
  • Präsentation von Ergebnissen auf Projektebene und auf internationalen Konferenzen und in Journalen
  • Die Möglichkeit zur Promotion ist gegeben

Unser Anforderungsprofil

  • Abgeschlossenes Masterstudium der Informatik, oder einem verwandten Fachgebiet
  • Fundierte Kenntnisse im Bereich Maschinelles Lernen, insbesondere Deep Learning
  • Idealerweise Erfahrung mit Neuromorphic Computing und/oder Event-Kameras
  • Gute Programmierkenntnisse in mindestens einer etablierten Programmiersprache
  • Kreativität, Selbstständigkeit, Teamgeist, und eine proaktive Einstellung
  • Gute deutsche und englische Sprachkenntnisse

Kontakt

Dr.-Ing. René Schuster (E-Mail: rschuste@rptu.de)

 

Researcher in the field of Neuromorphic Vision/Neuromorphic Sensing (m/f/d)

The Augmented Vision Lab is led by Prof. Dr. Didier Stricker and is part of the Department of Computer Science at RPTU and part of the German Research Center for Artificial Intelligence (DFKI). In ongoing research projects we are working on sustainable, energy efficient, and adaptive learning methods using event cameras and spiking neural networks (SNNs). For the next possible date we are looking for a new member for our team to answer research questions around the dynamic adaptation of neuromorphic sensors using neural networks.

Your responsibilites

  • Basic and application-oriented research in the field of neuromorphic computing and sensing
  • Simulation of event sensors in existing frameworks; implementation of adaptive sensor behavior
  • Work in a dynamic, international, and innovative working environment with excellent equipment
  • Training of neural networks with simulated event data
  • Presentation of results on project level and at international conferences and in journals
  • The possibility to do a PhD is given

Our job profile

  • Master’s degree in Computer Science, or a related field.
  • Sound knowledge in the field of machine learning, especially deep learning
  • Ideally experience with neuromorphic computing and/or event cameras
  • Good programming skills in at least one established programming language
  • Creativity, independence, team spirit, and a proactive attitude
  • Good German and English language skills

Contact

Dr.-Ing. René Schuster (E-Mail: rschuste@rptu.de)

Apply here.

Augmented Vision Department of DFKI led by Prof. Dr. Didier Stricker, offers a student assistant job (part-time) for the curious and passionate students, who want to develop themselves in Advanced Computer Vision.

Your Task

  • Utilize data augmentation techniques to create synthetic datasets
  • Apply state-of-the-art domain adaptation algorithms to analyze datasets for better adaption/generalization

Your Qualifications

  • Good knowledge of Python and Pytorch
  • Interest in Deep Learning and Computer Vision
  • Master or high semester Bachelor

Your Benefits

  • Skills in the methods of Domain Adaptation
  • Working in industry-oriented project with latest hardware
  • Practical experience in modern Deep Learning techniques
  • Flexible working times for better compatibility with studies

Apply latest by: 15.04.2024

Hours: by arrangement, 10h/week (30-40h/month)

Please feel free to contact us if you have any questions regarding this position:

yu.zhou@dfki.de

Room 2.03, DFKI-Kaiserslautern


Apply here.

GANs have made significant strides since their inception in 2014, demonstrating remarkable capabilities in generating realistic audio and video mixes, as well as complex geometries. However, a persistent challenge lies in ensuring the accuracy of the generated results. While the visual appeal of these results may be convincing, they often fail to faithfully represent the genuine geometric properties they aim to emulate. A pertinent example is the simulation of paper folding and crumpling, where maintaining the inherent geometric characteristics of the sheet (e.g., preventing stretching) is crucial. Although deterministic simulations of paper deformation have been developed, comprehending and replicating them often necessitate an extensive understanding of material physics, mathematics, and computer graphics. One potential approach to address this challenge involves harnessing the power of GANs or other network architectures, such as variational autoencoders, to analyze 3D geometries and generate highly accurate representations. However, several hurdles must be overcome, including operating effectively within the 3D space and establishing a robust methodology to evaluate the fidelity of the output geometries. An intriguing application of such methods for generating 3D geometries lies in their ability to rapidly generate synthetic data with exceptional accuracy. This synthesized data can then be employed for training purposes in various domains, offering a valuable resource for enhancing learning algorithms and expanding their applicability.

Tasks

  • State-of-the-art review on GANs and paper simulation
  • Generating paper geometries using simulation
  • Use GANs to generate geometries and designing accuracy and comparison methodology
  • Evaluating generated results from GANs and possible refinement strategies

Requirements

  • 3D computer vision, Computer graphics, Geometric modelling
  • Deep Learning (TensorFlow, PyTorch, Keras)
  • C++ (OpenGL, OpenCV), Python, C#

References

  • Goodfellow, I., Pouget-Abadie, 2020. Generative adversarial networks. Communications of the ACM, 63(11), pp.139-144.
  • Narain, R., 2013. Folding and crumpling adaptive sheets. ACM Transactions on Graphics (TOG), 32(4), pp.1-8.
  • Smith, E.J. ,2017, October. Improved adversarial systems for 3d object generation and reconstruction. In Conference on Robot Learning (pp. 87-96). PMLR.
  • Xiao, H., 2017. Fashion-mnist: a novel image dataset for benchmarking machine learning algorithms.
  • Das, S., 2019. Dewarpnet: Single-image document unwarping with stacked 3d and 2d regression networks.


Apply here.