Master-Seminar - Recent Highlights in Visual Data Analytics (IN2107)

Prof. Dr. Westermann

Time, Place: First meeting:
Wednesay, April 10., 2024 from 11:00-12:00 in room 02.13.010
Begin: The talks will be announced
Details:  
Prerequisites: Computer Graphics or Visualization, Proseminar
Registration: Please register via the matching system

Content

In this course, students will autonomously investigate recent research from the SIGGRAPH and IEEE VIS conference - the most renowned journals and conferences in the computer graphics and data visualization areas. Independent investigation for further reading, as well as a critical analysis and evaluation of the topic are required. Finally, the attendants have to present their results in a talk which should last 40-45 minutes. Talks will be given in English. Supplementary, a short written workout (approximately 4-6 pages) should be prepared.

Schedule

April 17, 2024 Deadline for paper preference submission
May 31, 2024 Deadline for report submission
First week of July Presentations (block seminar)

Papers

The paper list for this seminar can be found below. Please send us an email with your at least 4 ranked preferences by Wednesday, April 17th. 

We will assign papers based on these preferences, while remaining papers will be distributed randomly. The paper-matching results will be shared by mail.

Nb. Title      
1 A Comparative Visual Analytics Framework for Evaluating Evolutionary Processes in Multi-objective Optimization      
2 A Parallel Framework for Streaming Dimensionality Reduction      
3 Adaptive Sampling of 3D Spatial Correlations for Focus+Context Visualization      
4 ASTF: Visual Abstractions of Time-Varying Patterns in Radio Signals      
5 AttentionViz: A Global View of Transformer Attention      
6 Class-constrained t-SNE: Combining Data Features and Class Probabilities      
7 DendroMap: Visual Exploration of Large-Scale Image Datasets for Machine Learning with Treemaps      
8 Fast Compressed Segmentation Volumes for Scientific Visualization      
9 IDLat: An Importance-Driven Latent Generation Method for Scientific Data      
10 Interactive Focus+Context Rendering for Hexahedral Mesh Inspection      
11 Interactive Visual Cluster Analysis by Contrastive Dimensionality Reduction      
12 Interactive Volume Visualization via Multi-Resolution Hash Encoding based Neural Representation      
13 ManiVault: A Flexible and Extensible Visual Analytics Framework for High-Dimensional Data      
14 Parametric Dimension Reduction by Preserving Local Structure      
15 PC-Expo: A Metrics-Based Interactive Axes Reordering Method for Parallel Coordinate Displays      
16 Photon Field Networks for Dynamic Real-Time Volumetric Global Illumination      
17 PROWIS: A Visual Approach for Building, Managing, and Analyzing Weather Simulation Ensembles at Runtime      
18 PSRFlow: Probabilistic Super Resolution with Flow-Based Models for Scientific Data      
19 The Transform-and-Perform Framework: Explainable Deep Learning Beyond Classification      
20

Uncertainty-Aware Multidimensional Scaling

     
21 Uncertainty-Aware Principal Component Analysis      

 

The papers can be accesed through the institutional login on the publisher website (using TUM ID) or TUM library's eAccess.

Paper template

For your report, please use the VGTC Latex template. When generating the final PDF for submission, make sure to use the document class vgtc with option journal to obtain a report stylized as a final journal publication. The first page of the report should display the title, the author name and a brief abstract (before the start of the main text). A teaser figure on the first page is not required. To enable a standardized layouting of all reports, please use the layout as is. I.e. don't change margin sizes, font sizes, line spacings, header and footer sizes or related quantities, and avoid introducing additional white space beyond what is required by the template itself.