BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Multi-modal Image Processing: Data Models\, Algorithms\, and Appli
 cations - Miguel Rodrigues\, UCL 
DTSTART:20180525T120000Z
DTEND:20180525T130000Z
UID:TALK105652@talks.cam.ac.uk
CONTACT:Rachel Furner
DESCRIPTION:Many real-world data processing problems often involve heterog
 eneous images associated with different imaging modalities. These images a
 re often associated with the same phenomenon – sharing common attributes
  – so it is of interest to devise new mechanisms that can effectively le
 verage the availability of such multi-modal data in a number of data proce
 ssing tasks.\n\n \n\nThis talk proposes a multi-modal image processing fra
 mework based on joint sparse representations induced by coupled dictionary
  learning. In particular\, our framework can capture favorable structure s
 imilarities across different image modalities such as edges\, corners\, an
 d other elementary primitives in a learned sparse transform domain\, inste
 ad of the original pixel domain\, allowing us to develop new multimodal im
 age processing algorithms for a number of tasks.\n\n \n\nPractical experim
 ents with imaging data related to a number of applications – ranging fro
 m medical imaging\, art investigation\, and more – demonstrate that our 
 framework can lead to notable benefits in relation to other state-of-the-a
 rt approaches\, including deep learning algorithms.\n\n \n\nThis talk summ
 arizes joint work with various collaborators including Ingrid Daubechies (
 Duke U)\, Yonina Eldar (Technion)\, Lior Weizmann (Technion)\, Nikos Delig
 iannis (VUB)\, Bruno Cornellis (VUB)\, Pingfan Song (UCL)\, Joao Mota (Her
 iot Watt U)\, Pier Luigi Dragotti (Imperial College London)\, Xin Deng (Im
 perial College London)
LOCATION:MR5\, Centre for Mathematical Sciences
END:VEVENT
END:VCALENDAR