Recent News

Outstanding Editorial Board Member Award 2024

Sebastian Knorr was awarded the Outstanding Editorial Board Member Award 2024 from the IEEE Signal Processing Society for his service for the IEEE Transactions on Image Processing.

PicArrange now finds images with colors

New: PicArrange 3.4 Find Images by Text, Image, and Now, Color

🎨 Find images with similar colors
📍 Search photos by location
🧭 Filter images using GPS data

Tutorial on Visual Information Retrieval

The slides and Jupyter notebooks of our ECIR2025 tutorial are available.

Winner of the SISAP Indexing Challange

Our adaption of the Exploration Graph to low recall regimes won us the SISAP Indexing Challange.

About Us Visual Computing Group @ HTW Berlin

About Us

The Visual Computing Group is a collaboration of various researchers at the HTW Berlin who work in the fields of:

  • Machine Learning
  • Information Retrieval
  • Visual clustering & sorting
  • Computer Vision
  • Immersive Imaging
  • Neural Rendering
More about us

Recent Publications

Adaptive Segmentation-Based Initialization for Steered Mixture of Experts Image Regression

Yi-Hsin Li, Sebastian Knorr, Mårten Sjöström, Thomas Sikora
IEEE Transactions on Multimedia 2025

We introduce a novel adaptive segmentation-based initialization method targeted for optimizing Steered-Mixture-of Experts (SMoE) gating networks and Radial-Basis-Function (RBF) networks with steering kernels.

A Visual Quality of Experience Toolkit for Realistic Immersive Telepresence Applications

Manu Gond, Emin Zerman, Mohammadreza Shamshirgarha, Sebastian Knorr, and Mårten Sjöström
17th International Conference on Quality of Multimedia Experience (QoMEX) 2025

We introduce a visual quality of experience toolkit for realistic immersive telepresence applications

Real-Time View Synthesis with Multiplane Image Network using Multimodal Supervision

Manu Gond, Mohammadreza Shamshirgarha, Emin Zerman, Sebastian Knorr, and Mårten Sjöström
IEEE 27th International Workshop on Multimedia Signal Processing (MMSP) 2025

We introduce a framework that directly predicts multi-plane-image (MPI) parameters from a single RGB image for real-time view synthesis.

Geometry-aware video inpainting for joint headset occlusion removal and face reconstruction in social extended reality

Fatemeh Ghorbani Lohesara, Karen Eguiazarian, Sebastian Knorr
SPIE Journal of Electronic Imaging 2025

We introduce a geometry-aware learning-based framework to jointly remove HMD occlusions and reconstruct complete 3D facial geometry from RGB frames captured from a single viewpoint.