A Probabilistic Logic Programming Event Calculus

Anastasios Skarlatidis^1,2, Alexander Artikis¹, Jason Filippou^1,3 and Georgios Paliouras¹

¹Institute of Informatics and Telecommunications, NCSR “Demokritos”
²Department of Digital Systems, University of Piraeus
³University of Maryland, USA

Abstract:

We present a system for recognising human activity given a symbolic representation of video content. The input of our system is a set of time-stamped short-term activities (STA) detected on video frames. The output is a set of recognised long-term activities (LTA), which are pre-defined temporal combinations of STA. The constraints on the STA that, if satisfied, lead to the recognition of a LTA, have been expressed using a dialect of the Event Calculus. In order to handle the uncertainty that naturally occurs in human activity recognition, we adapted this dialect to a state-of-the-art probabilistic logic programming framework. We present a detailed evaluation and comparison of the crisp and probabilistic approaches through experimentation on a benchmark dataset of human surveillance videos.

Downloads

Paper (Draft)

The modified version of the CAVIAR dataset, including ProbLog-EC, the high-level event definitions and the event recognition results can be downloaded from here (Compressed archive ~0.5G; md5)

BibTex

@article{PLP-EC-TPLP15,
  author    = {Anastasios Skarlatidis and Alexander Artikis and Jason Filipou and Georgios Paliouras},
  title     = {{A Probabilistic Logic Programming Event Calculus}},
  journal   = {{TPLP}},
  volume    = {15},
  number    = {2},
  pages     = {213--245},
  year      = {2015},
  url       = {http://dx.doi.org/10.1017/S1471068413000690},
  doi       = {10.1017/S1471068413000690}
}