News: MICCAI 2018

Joint us on Wednesday, 19 September 2018 from 11:30 - 12:30, at this year's MICCAI for the presentation of our work on combining deep learning with non-linear statistical shape models for pancreas segmentation.

more info

News: CLIP 2018

Fraunhofer Singapore is co-organising the 7th Workshop on Clinical Image-based Procedures (CLIP 2018) held in conjunction with MICCAI in Granada, Spain, on 16 September. Join us for many exciting talks and keynotes about translational research in medical imaging.

more info

AI Image Analysis and Data Visualisation

AI Image Analysis

Understanding the content of images by identifying and classifying its constituents is an important task for many applications - be it in medical imaging where large amounts of 2D, 3D, and 4D images need to be analysed for diagnosis and operation planning every day or be it in automatic industrial production where images of components have to analysed in order to automatically detect production flaws. 

Artificial Intelligence technologies are nowadays fundamental for image analysis as they allow for the creation of more generic and more accurate image detection and classification algorithms and open the door to new applications that are promising to significantly boost productivity in many industry areas as well as to improve quality of life in general. The AI Image Analysis researchers at Fraunhofer Singapore have designed and applied Machine Learning methods to create intelligent systems and industry relevant applications for many different sectors.  

Automatic Model Based Segmentation of Anatomical Structures

Fraunhofer Singapore develops model based approaches for automatically segmenting anatomical structures like organs in medical images.   

Automatic Optical Inspection

    Our team develops active and deep learning algorithms for automatic image inspection. The methods are applied for production quality inspection in collaboration with the industry.

Micro X-Ray Analysis

Micro X-Ray is often applied for quality control of selected  samples in electrical curcuit and semiconductor device production. Fraunhofer Singapore has developed automatic quality control applications for analysis of Micro X-Ray images for industrial use where imperfections can be realiably detected.

2D Image Analysis

A lot of images we are dealing with today are still 2D, especially when it comes to historical data. Fraunhofer Singapore is developing AI methods to extract meaningful information in old, noisy and incomplete images, e.g. hand-drawn building plans.

Non-Linear Statistical Modelling

Using statistical modelling, complex information about shape and appearance of anatomical structures can be learnt, compressed and efficiently exploited for further image analysis. Our research focuses on building non-linear statistical models which are robust to errors in the training data.

Automatic object detection

We are creating deep learning algorithms to automatically detect structures and ananomalies in 2D, 3D, and 4D medical images. The methods are applied for Computer Aided Diagnosis and Decision Support.

Deformable Registration of Medical Images

Registration is the overlay of images through transformation. Using Fraunhofer Singapore's technology, images from the same patient acquired at different points in time or different imaging modalities can be precisely overlaid to judge the progress of a disease or the outcome of a surgery.

Video Analysis

Fraunhofer Singapore develops video analysis solutions based on deep learning in order to automatically categorise scenes according to their content. The methods can be applied for security, marketing, entertainment and others.

Data Visualisation

We are living in a world where the understanding of complex data is getting increasingly important in order to make informed decisions. A proper visualisation of this complex data is crucial in many application domains. The Data Visualisation team of Fraunhofer Singapore conducts research in the fundamentals of visualisation including computer graphics and real-time rendering and applies the outcomes in real-world applications. Examples are visualisation systems for Air Traffic Control, Building Information Management, Healthcare, Training and Education, Marketing, City Planning, IoT and Industrie 4.0 and many more.

Anatomical Visualisation

Fraunhofer Singapore conducts research in photo-realistic rendering and applies the results in order to create visualisations for education and anatomical training in cooperation with the Lee Kong Chian School of Medicine of NTU.

Group-based Learning Environments

    We are creating interactive learning spaces and platforms where students can interact with, discuss about, and share digital learning material, graphics, images, and other content.

Building Information Visualisation

    Fraunhofer Singapore has created a BIM visualisation platform that allows architects, decision makers, and construction companies to visualise large BIM models in real-time and to intuitively navigate in the 3D space through the use of 2D floorplan touch interaction.

Abstract Data Visualisation

    Fraunhofer Singapore has created a web-based platform that allows for the visualisation of complex data spaces in order to better understand relations and trends in the data. Furthermore, it allows for intuitive interaction using multi-touch interfaces.

3D City Model Rendering

Large 3D city models for city planning and simulation often consist of ten thousands of buildings. We are creating real-time rendering and geometry processing solutions that are able to realistically visualise complex 3D city models in real-time.

Visualisation of Cell Structures

    Biological processes within a cell are driven by structures that differ in size by many orders of magnitude. We are creating visualisation tools to better understand the inter-relations between these components on different scales.

Multi-Touch Interaction

Fraunhofer Singapore creates visual applications for multi-touch interfaces compatible with various touch surfaces, such as smartphones, tablet computer and more, using an in-house developed engine based on web technologies.

Future Air Traffic Control Visualisation

    Fraunhofer Singapore develops visualisation solutions for the future working place of air traffic controllers. Traditional 2D information as well as 3D information is merged in order to avoid air traffic conflicts and to allow for the handling of more aircraft in the future.

Sensor Data Visualisation

Fraunhofer Singapore has created an application to monitor and to visualise sensor data collected from various sources in real-time. Application examples are real-time monitoring of energy and water consumption in buildings and monitoring of IoT devices.

 

InfoLand

InfoLand is a web-based visualisation platform for touch-based interfaces like tablets, smart phones and touch tables. Check out the link below to download a trial version.

 

3D Stereoscopic Visualisation

    We are developing 3D Stereoscopic GPU Ray Casting to efficiently visualise 3D volume data, e.g. CT images, for application in Virtual Reality.

Jobs

You will find the latest job offers, PhD opportunities, and student theses projects on our career page.

Business opportunities

  • Contract Research and Development
  • Licensing of technologies
  • Consulting
  • Training and Services

Academic Collaborations

We are happy to explore academic and research collaborations and discuss potential topics with you.