direkt zum Inhalt springen

direkt zum Hauptnavigationsmenü

Sie sind hier

TU Berlin

Inhalt des Dokuments

Visual Processing

Motivation & Project Description

  • Integration of signal processing and computer vision experience in the human-computer interaction framework
  • Research in human action and interaction using multiple and heterogeneous camera systems 
  •  Three-fold research agenda:

    • Signal analysis
    • Low level vision
    • High level visual tasks (action recognition and 3D interaction)

Expected Outcome:
Novel multi-view video coding and signal enhancement; Depth map analysis and compression; Motion deblurring, image enhancement and 3D modelling; Robust feature extraction invariant to illumination changes that deals with camera motion and fusion of heterogeneous cameras; Person detection, body tracking and action recognition; User interface for interaction in 3D scenes.

Time Frame: 

05/2010 – 12/2012
T-labs Team Members:
Vladan Velisavljevic, Daniel Weinland, Michael Nicht, Rahul Swaminathan

Funding by:
Deutsche Telekom Laboratories

Zusatzinformationen / Extras


Schnellnavigation zur Seite über Nummerneingabe