1 ETISEO: Video Understanding Performance Evaluation Francois BREMOND, A.T. Nghiem, M. Thonnat, V. Valentin, R. Ma Orion project-team, INRIA Sophia Antipolis,

Slides:



Advertisements
Similar presentations
INRETS, Villeneuve dAscq, December 15 th -16 th 2005 Project Overview Video Understanding Evaluation David CHER R&D department R&D department SILOGIC S.A.,
Advertisements

By: Mani Baghaei Fard.  During recent years number of moving vehicles in roads and highways has been considerably increased.
ETISEO Project Corpus data - Video sequences contents - Silogic provider.
Road-Sign Detection and Recognition Based on Support Vector Machines Saturnino, Sergio et al. Yunjia Man ECG 782 Dr. Brendan.
ETISEO, Nice, May PETS International Workshops on Performance Evaluation of Tracking and Surveillance James Ferryman Computational Vision Group.
ETISEO Benoît GEORIS and François BREMOND ORION Team, INRIA Sophia Antipolis, France Lille, December th 2005.
PETS’05, Beijing, October 16 th 2005 ETISEO Project Ground Truth & Video annotation.
TAP-ET: TRANSLATION ADEQUACY AND PREFERENCE EVALUATION TOOL Mark Przybocki, Kay Peterson, Sébastien Bronsart May LREC 2008 Marrakech, Morocco.
MT Evaluation: Human Measures and Assessment Methods : Machine Translation Alon Lavie February 23, 2011.
Towards a Video Camera Network for Early Pest Detection in Greenhouses
Face Alignment with Part-Based Modeling
ELIS – Multimedia Lab Steven Verstockt T. Beji, B. Merci & R. Van de Walle RABOT2012 Presentation of a Multi-View Video Dataset of the Full-Scale (‘Rabot’)
INRETS, Villeneuve d’Ascq, December 15 th -16 th 2005 ETISEO Annotation rules Data structure Annotation tool and format Ground truth creation rules Reference.
Tracking Multiple Occluding People by Localizing on Multiple Scene Planes Professor :王聖智 教授 Student :周節.
Adviser : Ming-Yuan Shieh Student ID : M Student : Chung-Chieh Lien VIDEO OBJECT SEGMENTATION AND ITS SALIENT MOTION DETECTION USING ADAPTIVE BACKGROUND.
Efficient Moving Object Segmentation Algorithm Using Background Registration Technique Shao-Yi Chien, Shyh-Yih Ma, and Liang-Gee Chen, Fellow, IEEE Hsin-Hua.
Region-Level Motion- Based Background Modeling and Subtraction Using MRFs Shih-Shinh Huang Li-Chen Fu Pei-Yung Hsiao 2007 IEEE.
A Wrapper-Based Approach to Image Segmentation and Classification Michael E. Farmer, Member, IEEE, and Anil K. Jain, Fellow, IEEE.
CSE 291 Final Project: Adaptive Multi-Spectral Differencing Andrew Cosand UCSD CVRR.
Effective Gaussian mixture learning for video background subtraction Dar-Shyang Lee, Member, IEEE.
Background Subtraction for Urban Traffic Monitoring using Webcams Master Graduation Project Final Presentation Supervisor: Rein van den Boomgaard Mark.
Trinity College Dublin PixelGT: A new Ground Truth specification for video surveillance Dr. Kenneth Dawson-Howe, Graphics, Vision and Visualisation Group.
1 Real Time, Online Detection of Abandoned Objects in Public Areas Proceedings of the 2006 IEEE International Conference on Robotics and Automation Authors.
Jacinto C. Nascimento, Member, IEEE, and Jorge S. Marques
Image Subtraction for Real Time Moving Object Extraction Shahbe Mat Desa, Qussay A. Salih, CGIV’04.
INRIA, NICE, December 7 th -8 th 2006 Data Set and Ground Truth.
EE392J Final Project, March 20, Multiple Camera Object Tracking Helmy Eltoukhy and Khaled Salama.
Face Alignment Using Cascaded Boosted Regression Active Shape Models
Object detection, tracking and event recognition: the ETISEO experience Andrea Cavallaro Multimedia and Vision Lab Queen Mary, University of London
INRIA, Nice. December 7 th -8 th 2006 Evaluation protocol Evaluation process.
EADS DS / SDC LTIS Page 1 7 th CNES/DLR Workshop on Information Extraction and Scene Understanding for Meter Resolution Image – 29/03/07 - Oberpfaffenhofen.
1. Introduction Motion Segmentation The Affine Motion Model Contour Extraction & Shape Estimation Recursive Shape Estimation & Motion Estimation Occlusion.
Vision Surveillance Paul Scovanner.
Classroom Assessments Checklists, Rating Scales, and Rubrics
Usability testing. Goals & questions focus on how well users perform tasks with the product. – typical users – doing typical tasks. Comparison of products.
Video Event Recognition Algorithm Assessment Evaluation Workshop VERAAE ETISEO – NICE, May Dr. Sadiye Guler Sadiye Guler - Northrop Grumman.
ETISEO Benoît GEORIS, François BREMOND and Monique THONNAT ORION Team, INRIA Sophia Antipolis, France Nice, May th 2005.
ETISEO Evaluation Nice, May th 2005 Evaluation Cycles.
For ABA Importance of Individual Subjects Enables applied behavior analysts to discover and refine effective interventions for socially significant behaviors.
Project on Visual Monitoring of Human Behavior and Recognition of Human Behavior in Metro Stations with Video Understanding M. Thonnat Projet ORION INRIA.
Tracking and event recognition – the Etiseo experience Son Tran, Nagia Ghanem, David Harwood and Larry Davis UMIACS, University of Maryland.
AVITRACK Project FP INRIA WP1 - Apron Activity Model WP3 - Scene Tracking WP4 - Scene Understanding Brussels, January 17th 2006.
Efficient Visual Object Tracking with Online Nearest Neighbor Classifier Many slides adapt from Steve Gu.
ETISEO Benoît GEORIS and François BREMOND ORION Team, INRIA Sophia Antipolis, France Lille, December th 2005.
ETISEO Project Evaluation for video understanding Nice, May th 2005 Evaluation du Traitement et de l’Interprétation de Séquences vidEO.
VIP: Finding Important People in Images Clint Solomon Mathialagan Andrew C. Gallagher Dhruv Batra CVPR
Jack Pinches INFO410 & INFO350 S INFORMATION SCIENCE Computer Vision I.
Human Activity Recognition at Mid and Near Range Ram Nevatia University of Southern California Based on work of several collaborators: F. Lv, P. Natarajan,
By Naveen kumar Badam. Contents INTRODUCTION ARCHITECTURE OF THE PROPOSED MODEL MODULES INVOLVED IN THE MODEL FUTURE WORKS CONCLUSION.
INRETS, Villeneuve d’Ascq, December 15 th -16 th 2005 ETISEO Project Evaluation process.
Team Members Ming-Chun Chang Lungisa Matshoba Steven Preston Supervisors Dr James Gain Dr Patrick Marais.
CSSE463: Image Recognition Day 29 This week This week Today: Surveillance and finding motion vectors Today: Surveillance and finding motion vectors Tomorrow:
Suspicious Behavior in Outdoor Video Analysis - Challenges & Complexities Air Force Institute of Technology/ROME Air Force Research Lab Unclassified IED.
ETISEO François BREMOND ORION Team, INRIA Sophia Antipolis, France.
Tracking Groups of People for Video Surveillance Xinzhen(Elaine) Wang Advisor: Dr.Longin Latecki.
Lecture 04 Edge Detection Lecture 04 Edge Detection Mata kuliah: T Computer Vision Tahun: 2010.
PETS’05, Beijing, October 16 th 2005 ETISEO Project Video Providers Corpus Data Video contents.
Fast Human Detection in Crowded Scenes by Contour Integration and Local Shape Estimation Csaba Beleznai, Horst Bischof Computer Vision and Pattern Recognition,
Motion tracking TEAM D, Project 11: Laura Gui - Timisoara Calin Garboni - Timisoara Peter Horvath - Szeged Peter Kovacs - Debrecen.
Video object segmentation and its salient motion detection using adaptive background generation Kim, T.K.; Im, J.H.; Paik, J.K.;  Electronics Letters 
Real-Time Soft Shadows with Adaptive Light Source Sampling
Contents Team introduction Project Introduction Applicability
Traffic Sign Recognition Using Discriminative Local Features Andrzej Ruta, Yongmin Li, Xiaohui Liu School of Information Systems, Computing and Mathematics.
Tasks processing Strong participation . Large results comparison on
Compositional Human Pose Regression
Scene Understanding Francois BREMOND
Vehicle Segmentation and Tracking in the Presence of Occlusions
SILOGIC S.A. , Toulouse, France
Introduction Computer vision is the analysis of digital images
Evaluation of UMD Object Tracking in Video
Presentation transcript:

1 ETISEO: Video Understanding Performance Evaluation Francois BREMOND, A.T. Nghiem, M. Thonnat, V. Valentin, R. Ma Orion project-team, INRIA Sophia Antipolis, FRANCE

2 Outline Introduction ETISEO Project Video Data ETISEO Results Metric Analysis ETISEO General Conclusion

3 Introduction There are many evaluation initiatives with different objectives Individual works projects: CAVIAR, ILids, VACE, CLEAR, CANTATA,… Workshops: PETS, VS, AVSS (CREDS),… Issues: Not standard annotation (ground truth) Lack of analysis of Video Data which specific video processing problems a sequence contains how difficult these problems are Lack of analysis of metrics Numbers, base-line algorithm

4 ETISEO Project 2 years duration, from January 2005 to December 2006 To evaluate vision techniques for video surveillance applications. Goals: Unbiased and transparent evaluation protocol (no funding) Large involvement (32 international teams) Meaningful evaluation provide the strengths and weaknesses of metrics to help developers to detect specific shortcomings depending on –scene type (apron, building entrance etc.) –video processing problem (shadows, illumination change etc.) –difficulty level (e.g. strong or weak shadows)

5 Approach: 3 critical evaluation concepts Ground truth definition Rich and up to the event level Give clear and precise instructions to the annotator E.g., annotate both visible and occluded part of objects Selection of test video sequences Follow a specified characterization of problems Study one problem at a time, several levels of difficulty Metric definition various metrics for each video processing task Performance indicators: sensitivity, precision and F-score. A flexible and automatic evaluation tool, a visualization tool. ETISEO Project

6 4 Companies: - Barco - Capvidia NV - VIGITEC SA/NV - Robert Bosch GmbH 12 Academics: - Lab. LASL University ULCO Calais - Nizhny Novgorod State University - Queen Mary, University of London - Queensland University of Technology - INRIA-ORION - University of Southern California - Université Paris Dauphine - University of Central Florida - University of Illinois at Urbana-Champaign - University of Maryland - University of Reading - University of Udine ETISEO Project : Large participation (16 active international teams)

7 ETISEO : Video Data Large annotated data set 85 video clips with GT, organized into scene types : apron, building entrance, corridor, road, metro station, video processing problems : noise, shadow, crowd, … sensor types : one\multi-views, visible\IR, compression…

8 : Airport Video Data : Airport Silogic Toulouse – France Apron Multi-view

9 INRETS-LEOST Villeneuve d’Ascq – France : INRETS Video Data : INRETS Building Entrance Car Park Light Changes

10 : CEA Video Data : CEA Street Corridor Video Type & Quality

11 : RATP Video Data : RATP Subway People Density

12 Detection of physical objects ETISEO : Results 16 Teams Detection rate Evaluation on 6 videos

13 Tracking of physical objects ETISEO : Results Detection rate Teams

14 ETISEO: Results Good performance comparison per video: automatic, reliable, consistent metrics: 16 participants: 8 teams achieved high quality results 9 teams performed event recognition 10 teams produced results on all priority sequences Best algorithms: combine moving regions and local descriptors A few limitations: Algorithm results depend on time processing (RT), manpower (parameter tuning), previous similar experiences, learning stage required or not…: questionnaire Lack of understanding of the evaluation rules (output XML, time-stamp, ground truth, number of processed videos, frame rate, start frame…) Video subjectivity: background, masks, GT (static, occluded, far, portable, contextual object, event) Many metrics and evaluation parameters Just evaluation numbers, no base-line algorithm Need of two other analyses: 1. Metric Analysis: define for each task: Main metrics: discriminate and meaningful Complementary metrics: provide additional information 2. Video Data Analysis: impact of videos on evaluation define a flexible evaluation tool to adapt GT wrt videos

15 Metric Analysis : Object detection task Main metric: Number of objects Evaluate the number of detected objects matching reference objects using bounding box Unbiased towards large, homogenous objects Difficult to evaluate object detection quality Complementary metric: Object area Evaluate the number of pixels in reference data that have been detected Evaluate the object detection quality Biased toward large, homogenous objects

16 Metric Analysis : Example (1) Sequence ETI-VS2-BE-19-C1 has one big object (car) and several small and weakly contrasted objects (people)‏ Algorithm 9 correctly detects more objects than algorithm 13 (metric: Number of objects)‏ F-Score Algorithm F-Score Algorithm Performance results using the metric “number of objects”

17 Metric Analysis : Example (2) Using metric Object area, biased toward big object (car): algorithm 13 cannot detect some small objects (people), algorithm 9 has detected difficult objects at low precision. Metric Object area is still useful: it differentiates algorithms 1 and 9: both are good at detecting objects but algorithm 1 is more precise F-Score Algorithm F-Score Algorithm Performance results using the metric “object area”

18 Metric Analysis : Advantages & Limitations Advantages : various metrics for every video processing task. analysis of the metric strengths and weaknesses and how to use them. insight into video analysis algorithms: for example, shadows, merge Still some limitations : Evaluation results are useful for developers but not for end-users. Ok, not a competition nor benchmarking But difficult to judge if one algorithm is good enough for a particular application, or type of videos.

19 ETISEO limitations: Generalization of evaluation results is subjective : comparing tested and new videos Selection of videos according to difficulty levels is subjective Videos have only qualitative scene description: eg. strong or weak shadow Two annotators may assign 2 different difficulty levels One video may contain several video processing problems at many difficulty levels The global difficulty level is not sufficient to identify algorithm's specific problems for improvement ETISEO : Video Data Analysis

20 Objectives of Video Data Analysis : Study dependencies between videos and video processing problems to Characterize videos with objective difficulty levels Determine algorithms capacity in solving one video processing problem. Approach: To treat each video processing problem separately Define a measure to compute difficulty levels of videos (or other input data) Select videos containing only the current problems at various difficulty levels For each algorithm, determine the highest difficulty level for which this algorithm still has acceptable performance. Approach validation : applied to two problems Detection of weakly contrasted objects Detection of objects mixed with shadows Video Data Analysis

21 Video Data Analysis : Detection of weakly contrasted objects Video processing problem definition : the lower the object contrast, the worse the object detection performance For one algorithm, determine the lowest object contrast for which this algorithm has an acceptable performance Issue: one blob may contain many regions at several contrast levels

22 Video Data Analysis : conclusion Achievements: An evaluation approach to generalise evaluation results. Implementation of this approach for 2 problems. Limitations: Need to validate this approach for more problems. Works well if the video contains only one problem. If not, detects the upper bound of algorithm capacity. The difference between the upper bound and the real performance may be significant if: The test video contains several video processing problems The same set of parameters is tuned differently to adapt to several dependent problems

23 General Conclusion Achievements: Good performance comparison per video: automatic, reliable, consistent metrics. Emphasis on gaining insight into video analysis algorithms (shadows, occlusion,..) A few limitations: Data and rule subjectivity: background, masks, ground truth,… Partial solutions for Metric and Video dependencies Future improvements: flexible evaluation tool Given a video processing problem: Selection of metrics Selection of reference videos Selection of Ground Truth : filters for reference data, sparse GT for long videos ETISEO’s video dataset and automatic evaluation tools are publicly available for research purposes:

24 At each contrast level, the algorithm performance is x/m x: number of blobs containing current contrast level detected by a given algorithm m: number of all blobs containing current contrast level Algorithm capacity: the lowest contrast level for which algorithm performance is bigger than a given threshold Video Data Analysis : Detection of weakly contrasted objects

25 Video Data Analysis : Detection of weakly contrasted objects Error rate threshold to determine algorithm capacity: 0.5