AT&T Research at TRECVID 2013: Surveillance Event Detection

نویسندگان

  • Xiaodong Yang
  • Zhu Liu
  • Eric Zavesky
  • David Gibbon
  • Behzad Shahraray
  • YingLi Tian
چکیده

This paper presents a general surveillance event detection system that participated in the interactive Surveillance Event Detection (iSED) task of TRECVID 2013. In the proposed system, a set of spatio-temporal features including Space-Time Interest Points (STIP) and Dense Trajectories are extracted, and a sliding temporal window is employed as the detection unit. Fisher Vector is used to encode low-level features as the representation of each sliding window. Both feature-level and decision-level fusions are used to combine multiple features. In order to deal with the highly imbalanced nature of surveillance data, the system performs detections using the CascadeSVMs algorithm according to each specific event and camera view. Two different interactive environments are evaluated, one focuses on high throughput and the other includes related result expansion. In the primary run evaluations, our system ranks the top in 2 out of 7 event detection tasks.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

NHK STRL at TRECVID 2009: Surveillance Event Detection and High-Level Feature Extraction

NHK Science and Technology Research Laboratories participated in two tasks at TRECVID 2009: surveillance event detection task and high level feature extraction task. For surveillance event detection tasks, we targeted four events: " P e r s o n R u n s " , " P e o p l e M e e t " , " O b j e c t P u t " , a n d "OpposingFlow". The proposed method detects human regions using HOG descriptor and S...

متن کامل

Toshiba at TRECVID 2008: Surveillance Event Detection Task

In this paper, we describe the Toshiba event detection system for TRECVID surveillance event detection task [1] that detects three TRECVID required events (E05:PersonRuns, E19:ElevatorNoEntry, and E20:OpposingFlow). Our system (“Toshiba_1 p-cohog_1”) consists of four components: (1) robust change detection based on the combination of pixel intensity histogram, PTESC (Peripheral TErnary Sign Cor...

متن کامل

PKU-IDM @ TRECVid 2010: Pair-Wise Event Detection in Surveillance Video

In this paper, we describe our system for the surveillance events detection task in TRECVid 2010. We focused on pair-wise events (e.g., PeopleMeet, PeopleSplitUp, Embrace) that need to explore the relationship between two active persons. For our team had participated in the TRECVid SED task in 2009, we developed the system based on the old one. The improvements are three-fold. First, we refined...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2013