[Forschungsseminar-BSV] Forschungsseminar Computergrafik, Bildverarbeitung und Visualisierung

[Vanessa Kretzschmar]

Dear all,

I want to announce a second presentation for upcoming Wednesday (April 
the 28th, 2021):

the talk will be given by

     Christian Blecha

and is entitled:

     'Visual Analysis of the Relation Between Stiffness Tensor and the 
Cauchy-Green Tensor'

Abstract:

     Stress and strain tensors, two well-known quantities in mechanical
     engineering, are linked through a fourth-order stiffness tensor, which
     is not considered by many visualizations due to its complexity. 
Considering
     an orthotropic material, the tensor naturally decomposes into nine 
known
     material properties. We used fiber surfaces to analyze a data set
     representing a biological tissue. A sphere is pushed into the material
     to confirm the mathematical link as well as the possibility to extract
     highly deformed regions even if only the stiffness tensor is 
available.

Yours sincerely
Vanessa Kretzschmar

On 4/19/2021 1:21 PM, Vanessa Kretzschmar wrote:
> Achtung:
>  - anderer Raum
>
>
> E I N L A D U N G
>
> ======================================================================
>
> zum Forschungsseminar 'Computergrafik, Bildverarbeitung und 
> Visualisierung'
>
>     am Mittwoch, den 28. April 2021, 13:15 Uhr,
>     über eine Webkonferenz. 
> (https://conf.fmi.uni-leipzig.de/b/van-bwb-jil-vk2)
>
> ======================================================================
>
> Wir hören einen Vortrag von
>
>     Lars Nieradzik
>
> mit dem Titel:
>
>     'Uncertainty estimation in neural networks with application to 
> medical image segmentation'
>
> zum Inhalt:
>
>     Estimating uncertainty has become an important area of research in 
> machine
>     learning. In this thesis, several methods for measuring and improving
>     probabilistic estimates are evaluated on common image datasets. We 
> show
>     that calibration metrics are inferior to proper scoring rules in 
> medical
>     image segmentation and general classification tasks. Furthermore, two
>     new metrics for segmentation are proposed to measure the 
> uncertainty at
>     foreground pixels (e.g. at lesions). Taking the task of finding 
> ischemic
>     stroke lesions in CT scans as an example, we build a neural 
> network that
>     displays probabilistic estimates for injured tissue. The network 
> reaches
>     new state-of-the-art results on a dataset provided by the Uniklinikum
>     Leipzig. A visual inspection of the predictions reveals that the
>     time-to-maximum of the residue function (Tmax) is highly 
> correlated with
>     the predictions that the network makes. We recommend discarding 
> Tmax and
>     using raw spatio-temporal data for learning better feature 
> representations
>     of stroke lesions.
>
> ======================================================================
>
> Alle Interessierten sind im Namen von Professor Dr. Scheuermann herzlich
> eingeladen.
>
> Mit freundlichen Grüßen
> Vanessa Kretzschmar