Sharing Lifelog Experiences ODCSSS 2008 Paula Meehan DCU Prof. Noel O Connor, Daragh Byrne
Contents Introduction Background Research Event Segmentation Matching Events (1) Matching Events (2) Progress Model
Introduction What is a Lifelog? Connecting different user’s Lifelog Collections.
Background Research CDVP (Centre for Digital Video Processing) Event Segmentation Content and Context in Multimedia Semantics MPEG feature extraction
Event Segmentation A day’s SenseCam images (3,000 – 4,000) Multiple Events Finishing work in the lab At the bus stop Chatting at Skylon Hotel lobby Moving to a room Tea time On the way back home Event Segmentation A day’s SenseCam images (3,000 – 4,000) Multiple Events Finishing work in the lab At the bus stop Chatting at Skylon Hotel lobby Moving to a room Tea time On the way back home Event Segmentation Share Events
Event-Segmented image sets User 1 User 2 Compare Event-Event similarity between  Users Event’s attended by both Users Matching Events (1) ... : Similarity matrix
User 2  User 1  Matching Events (2) Similarity Score : Scalable Colour Colour Structure Colour Layout  Colour Moments Edge Histogram Homogeneous Texture Extract MPEG-7 descriptors for this image  Scalable Colour Colour Structure Colour Layout  Colour Moments Edge Histogram Homogeneous Texture Extract MPEG-7 descriptors for this image  :
Progress Presently Parsing ‘sensor.txt’ and ‘image.xml’ files Future Work Plot data and compare Extract MPEG-7 Descriptor features of images Match Events of Users Solution for timestamps of events which aren’t synchronised
Model My Events User 1’s Events My Shared Event User 1’s Shared Event Shared Event
Thank you

Sharing Lifelog Experience (Midterm)

  • 1.
    Sharing Lifelog ExperiencesODCSSS 2008 Paula Meehan DCU Prof. Noel O Connor, Daragh Byrne
  • 2.
    Contents Introduction BackgroundResearch Event Segmentation Matching Events (1) Matching Events (2) Progress Model
  • 3.
    Introduction What isa Lifelog? Connecting different user’s Lifelog Collections.
  • 4.
    Background Research CDVP(Centre for Digital Video Processing) Event Segmentation Content and Context in Multimedia Semantics MPEG feature extraction
  • 5.
    Event Segmentation Aday’s SenseCam images (3,000 – 4,000) Multiple Events Finishing work in the lab At the bus stop Chatting at Skylon Hotel lobby Moving to a room Tea time On the way back home Event Segmentation A day’s SenseCam images (3,000 – 4,000) Multiple Events Finishing work in the lab At the bus stop Chatting at Skylon Hotel lobby Moving to a room Tea time On the way back home Event Segmentation Share Events
  • 6.
    Event-Segmented image setsUser 1 User 2 Compare Event-Event similarity between Users Event’s attended by both Users Matching Events (1) ... : Similarity matrix
  • 7.
    User 2 User 1 Matching Events (2) Similarity Score : Scalable Colour Colour Structure Colour Layout Colour Moments Edge Histogram Homogeneous Texture Extract MPEG-7 descriptors for this image Scalable Colour Colour Structure Colour Layout Colour Moments Edge Histogram Homogeneous Texture Extract MPEG-7 descriptors for this image :
  • 8.
    Progress Presently Parsing‘sensor.txt’ and ‘image.xml’ files Future Work Plot data and compare Extract MPEG-7 Descriptor features of images Match Events of Users Solution for timestamps of events which aren’t synchronised
  • 9.
    Model My EventsUser 1’s Events My Shared Event User 1’s Shared Event Shared Event
  • 10.