think-cell Software on LinkedIn: VMV at TU Braunschweig, Germany 2023 | think-cell (2024)

Allresearchers studying multi-robot systemsare invited to complete a short survey about the research resources needed to accelerate research in this area:https://lnkd.in/eaAZxnf6With our academic partners, the University of Maryland plans to lead the development of the Infrastructure for Collaborative Autonomy Testing (ICAT), which will support computer science and engineering researchers who are studying cooperative multi-robot systems (MRS). In particular, ICAT will be a collection of software and hardware resources that reduce the barriers to MRS research, reduce the time and effort to deploy MRS, and support replicability and reproducibility.Planned resources include simulation models, standardized MRS missions, tools to setup, run, and collect data from simulation models, templates for documenting results, and a network of robot test facilities.This information-gathering survey will help us understand and support the specific needs of this research community.The survey does not ask for any personal data.

9

Another exciting news! As an essential part of my dissertation, "A Global Correction Framework for Camera Registration in Video See-Through Augmented Reality Systems," is officially online in The ASME Journal of Computing and Information Science in Engineering (JCISE)! 📚 Check it out here: https://lnkd.in/dVzuSaJwOPEN the floor for any questions, discussions, and citations about this promising topic.Let's engage in this insightful research to propel the XR world forward!

26

Navigating the complexities of scientific research often involves juggling large data sets, multiple tools, and specialized software–especially in the realm of high performance computing (HPC). In an environment where the lack of reproducibility in scientific studies is a growing concern, tools that enhance accuracy and consistency are invaluable. In the Ebbert Lab, a part of […] #Engineering #Robotics #Recruiters

Another one of my work in scientific computing showcasing "Fast GPU-accelerated scheme for the simulations of extensible capusles" is out on arXiv now. Here, we devised a novel numercial scheme to simulate capsules in Stokesian flows. Our scheme has fourth order convergence and offers locally adaptable parameterization of surfaces. Moreover, with FMM acceleration its complexity can be reduced to O(N) work complexity which is a tremendous improvement over the spherical harmonics based spectral scheme which has O(N^2 logN ) work complexity. arXiv link: https://lnkd.in/eHStBU_U

31

2 Comments

🔓 New paper unlocked! Our work, "A Machine Learning-Oriented Survey on Tiny Machine Learning", has been accepted by IEEE Access!🔗 Paper link: https://lnkd.in/duuVswpt.This paper proposes a taxonomy that covers the learning panorama under the #TinyML lens, examining the different families of model optimization, design, and state-of-the-art learning techniques in detail.Furthermore, you will find information like the popularity over the years in terms of research contributions, the distribution of application domains highlighting the most common TinyML use cases, the TinyML off-the-shelf hardware, and their usage in current literature.A big thanks to all the other co-authors: Federico Cunico, Dong Seon Cheng, Franco Fummi, and Marco Cristani 💪

31

[CAV 2023 - Conference]🤝 Irif is co-organizing the 35th International Conference on Computer Aided Verification (CAV) which will be held this year in Paris. CAV 2023 is the 35th in a series dedicated to the advancement of the theory and practice of computer-aided formal analysis methods for hardware and software systems. The conference covers the spectrum from theoretical results to concrete applications, with an emphasis on practical verification tools and the algorithms and techniques that are needed for their implementation. CAV considers it vital to continue spurring advances in hardware and software verification while expanding to new domains such as machine learning, autonomous systems, and computer security.📅 July 17th-22nd, 2023 (the first two days will be dedicated to affiliated workshops)📌 MAISON DE LA CHIMIE, 28 rue Saint Dominique 75007 Paris🔗 To know more about the program, the keynotes and to register : http://www.i-cav.org/2023/ #software #algorithms #computersecurity #hardware #machinelearning

4

So many people knowing Newton and Einstein, but not Claude Shannon. With the importance of digital communication and the future of AI it is time for this to change. Looking at the introduction of his 1948 paper “A Mathematical Theory of Communications” this paragraph may in hindsight seem straightforward (as so often is the case), but this way of thinking is what allowed him to develop the foundations of information theory:"The fundamental problem of communication is that of reproducing at one point either exactly or approximately a message selected at another point. Frequently the messages have meaning; that is they refer to or are correlated according to some system with certain physical or conceptual entities. These semantic aspects of communication are irrelevant to the engineering problem. The significant aspect is that the actual message is one selected from a set of possible messages. The system must be designed to operate for each possible selection, not just the one which will actually be chosen"

12

1 Comment

I’ll happily admit I get my name on this paper mostly for identifying that simulating many varied aggregate structures and the drag they experience under selected conditions allows exploration of different 2D aggregate size measures and drag models, to then choose which gives the smallest errors to derive aggregate density. The computational aspects are well beyond me, so fortunately I could rely on Ismael Mola and Michael Small (with CSIRO’s computational resources) to maximise outcomes. Real aggregate structures are complex and not that fractal, so reducing errors from 2D imaging allows more insightful comparisons. This work owes its origins to groups from the 1980s that include John Farrow & Len Warren (CSIRO) and Jerzy Gańczarczyk & Da-Hong Li (University of Toronto) – what they achieved back then without modern cameras and advanced computing still amazes me.https://lnkd.in/gdRSMQNt

48

1 Comment

🚀 Exciting News! Our paper, titled "WALDO: Future Video Synthesis using Object Layer Decomposition and Parametric Flow Prediction," will be presented at the International Conference on Computer Vision (ICCV) in Paris next week! 🎉This work was done at Inria in collaboration with Cordelia Schmid and Jean Ponce.🔮 Future video synthesis is no easy task, as it involves predicting new frames based on past ones. It requires a deep understanding of the world, encompassing everything from physics to social behaviors. Tackling this challenge at high-definition and for extended video sequences calls for memory- and compute-intensive methods.🎥 Enter WALDO, an innovative approach that breaks down video synthesis into manageable subtasks:1️⃣ Context Understanding: We decompose past frames into distinct layers, producing masks that segment objects and control points that account for their deformations.2️⃣ Future Prediction: We extrapolate the trajectory of these control points into the future.3️⃣ Frame Synthesis: We use the dense motion induced by these control points to warp past frames into the future and inpaint missing regions.Our strategy sets a new state of the art across various benchmarks featuring rigid and non-rigid motions, and produces realistic videos even in the face of uncertainty. Discover more about WALDO:🌐 Webpage: https://lnkd.in/ePMyk6_U📄 Paper: https://lnkd.in/exyg4Q_X💻 Code: https://lnkd.in/eNC4y5wmJoin us for the poster presentation at ICCV Paris:📅 Date: October 6th🕝 Time: 2:30-4:30 PM

73

1 Comment

So many outstanding UCF faculty are unleashing the potential of modeling and simulation. Dr. Amon shares her unique research and teaching approach to problem-solving, social networks, human-plus-AI, human-within-VR and other human-computer and social interactions. https://ow.ly/URbx50PWEqV

32