Perception system for a household assistant robot
We were working on the perception system of the robot. By developing object pose estimation and tracking algorithms we allowed Aeolus to use robot's arm to perform various operations on objects e.g. grabbing a can of Coca-Cola
Our team assisted Aeolus from the beginning. We assisted in establishing key components of the system, we designed build environment & test structure for whole organization
Robot was presented on CES 2018
A novel segmentation algorithm dedicated to budding yeast
Automatic tracking of cells in time-lapse microscopy is required to investigate a multitude of biological questions. To limit manipulations during cell line preparation and phototoxicity during imaging, brightfield imaging is often considered. Since the segmentation and tracking of cells in brightfield images is considered to be a difficult and complex task, a number of software solutions have been already developed.
CellStar is one of such algorithms. It is optimized to segment and track images of budding yeast cells growing in monolayer (e.g. images from microfluidic chambers), however the algorithm can be also used to track other round objects (in brightfield as well as fluorescent images).
A specie annotation tool for Deep Learning
Taxonify is a platform that expedites the process of species annotation for automated classification. It supports image browsing and annotation in a way that is optimal for taxonomists and their annotation workflow. Taxonify is free and open source - you can download, modify it and use it according to your needs!
It's currently used in the project Aquascope for annotation of phytoplankton and zooplankton images coming from automated underwater imaging cameras in lake Greinfensee (Switzerland). The software was funded by the Swiss National Science Foundation project 310030L_182124 (Aquascope). You can check out more about Aquascope project on their website - www.aquascope.eawag.ch
Microscoft's Hololens tool for médecine
Essential Vision is application for visualization of time series 3D models and simulation results run on them. Application allows to share view with multiple users which makes it suitable for learning classes e.g. morphology class. It is dedicated for Hololens headset. Holoseries is free and open source - you can download, modify it and use it according to your needs!
It is a shared project between ETH Zurich and EXCITE.
The application enables you to remove cranes from any photo. Check out the app.
EXAMPLES OF IMAGE ANALYSIS CAPABILITIES - LIFE SCIENCE
Other examples of our capabilities
Colony counting based on a Petri dish image
Segmentation of a central venous catheter removed from an asymptomatic patient
Segmentation of methicillin-resistant Staphylococcus aureus bacteria
Monoclonal Antibodies Bound to HeLa Cells
Histopathology of a keratocystic odontogenic tumour. Hematoxylin and Eosin (H&E)
Histopathology of a glomus tumor tumour. Hematoxylin and Eosin (H&E)
Blood vessel segmentatation in retinal scan images
Bood cells segmentation
EXAMPLES OF IMAGE ANALYSIS CAPABILITIES - Other
Other examples of our capabilities
Carbon nanofibers obtained by electrospinning from a polymer percurson
Scanning an electron micrograph (SEM) of PLGA nanoparticles
A original image captured from NASA's World Wind software for water detection
Jellyfish detection at Monterey Bay, CA
Aircraft detection at the Newark Liberty International Airport
Check out some of our projects (minus some details which have been removed due to NDAs)
Microscopy, Image Processing, R&D
From our experience and research, we designed and developed an algorithm for nuclei segmentation in microscopy images which is now a key feature of Leica software. The following paper presents additional concepts not covered by the NDA.
A Robust Algorithm for Segmenting and Tracking Clustered Cells in Time-Lapse Fluorescent Microscopy (IEEE Journal of Biomedical and Health Informatics, 2013).
European Space Agency
Image Processing, Cloud, R&D
Space debris refers to the increasing number of man-made objects orbiting Earth, such as obsolete satellites. As of July 2013, about 670,000 small items (1-10 cm in size) and 29,000 larger items were estimated to be in orbit. To help resolve this problem, we developed scalable image analysis algorithms that enable debris detection in images collected by telescopes.
Computer Vision, Machine Learning, R&D
We supported Aeolus Robotics by leading the development of a vision system that integrated with a revolutionary product they were developing. This system required competency in image analysis and computer vision, which our team provided. In addition, our teams also used recent Artificial Intelligence and Machine Learning advancements to extend standard approaches.
MicroscopeIT spreads its ideas to dedicated spinoffs & other business entities
LifeFlow specializes in cardiac diagnostics. Our team, which consists of bioengineers, cardiologists, and ICT professionals, has created an innovative, noninvasive method for identifying atherosclerosis via algorithms used in computational fluid dynamics (CFD).
We supported LifeFlow from its inception by providing consultation and an early prototype of a diagnostic tool.
MicroscopeIT's CSO is the co-founder of LifeFlow.
Deep Learning, Cloud Computing
Histopixel specializes in histopathological diagnostics. We apply state-of-the-art machine learning techniques to improve and accelerate the diagnosis of cancer through histopathological images.
We created a unique software for Histopixel that enables visualizing, annotating, and processing histopathological images. Cloud computations support bot deep learning techniques and the morphological analysis of data. The algorithm we proposed also reaches true positive rates of >84%, which is slightly above the level achieved by humans and was
a state-of-the-art accomplishment in 2016.
Histopixel is a spinoff of MicroscopeIT.
Game Studio & Design Shop
An indie game studio created out of love for cats and computer games. CAG specializes in games featuring an engaging story and unique comic-inspired graphics. To develop games the studio uses its proprietary open-source 3D/2D Castle Game Engine.
MicroscopeIT's CEO is a co-founder of Catastrophe Games.