Fraunhofer HHI at TRECVID 2004: Shot Boundary Detection System
This paper describes the shot boundary detection and determination system developed at the Fraunhofer Institute for Telecommunications, Heinrich-HertzInstitut, used for the evaluation at TRECVID 2004. The system detects and determines the position of hard cuts, dissolves, fades, and wipes. It is very fast and has proved to have a very good detection performance. As input for our system, we use luminance pixel values of sub-sampled video data. The hard cut detector uses pixel and edge differences with an adaptive thresholding scheme. Flash detection and slow motion detection lower the false positive rate. Dissolve and fade detection is done with edge energy statistics, pixel and histogram differences, and a linearity measure. Wipe detection works with an evenness factor and double Hough transform. The difference between the submitted runs is basically only different threshold settings in the detectors, resulting in different recall and precision values.
CLIPS at TRECVID : Shot Boundary Detection and Feature Detection
This paper presents the systems used by CLIPS-IMAG to perform the Shot Boundary Detection (SBD) task and the Feature Extraction (FE) task of the TRECvid workshop. Results obtained for the 2003 evaluation are presented. The CLIPS SBD system based on image difference with motion compensation and direct dissolve detection was second among 14 systems. This system gives control of the silence to noise ratio over a wide range of values and for an equal value of noise and silence (or recall and precision), the value is 12 % for all types of transitions. Detection of person X from speaker recognition alone was deceiving due to the small number of shots containing person X in the overall test collection (about 1/700) and the even small number in which person X was actually speaking (about 1/6000). Detection of person X from speech transcription performed much better but was still lower than other systems using also the image track for the detection.
Shot Boundary Determination on MPEC Compressed Domain and Story Segmentation Experiments for TRECVID 2003
1. Briefly, what approach or combination of approaches did you test in each of your submitted runs? 1_kddi_ss_base1_5: “Baseline” method based on SVM, which discriminates shots that contain story boundaries. 1_kddi_ss_c+k1_4: Baseline + section-specialized segmentation (SS-S). 1_kddi_ss_all1_3: Baseline + SS-S + anchor shot segmentation (ASS) based on audio classification results 1_kddi_ss_all1_pfil_1: Baseline + SS-S + ASS and post-filtering (PF) based on audio classification results 1_kddi_ss_all2_pfil_2: Extended baseline + SS-S + ASS + PF based on audio classification results. 1_kddi_ss_all1nsp07_pfil_6: Baseline + SS-S + ASS + PF by HMM-based non-speech detection. 1_kddi_ss_all2nsp07_pfil_7: Extended baseline + SSS + ASS + PF by HMM-based non-speech detection. 2_kddi_ss2_all1_pfil_8: Baseline + SS-S + ASS and PF based on "speech segment" information from LIMSI ASR results[1]. 2_kddi_ss2_all2_pfil_9: Extended baseline + SS-S + ASS and PF based on "speech segment" information from LIMSI ASR results. 3_kddi_ss3_10: Naive TextTiling based story segmentation based on LIMSI ASR data.
neural network sensor network wireless sensor network wireless sensor deep learning comparative study base station information retrieval feature extraction sensor node programming language cellular network random field digital video number theory rate control network lifetime river basin hyperspectral imaging distributed algorithm chemical reaction carnegie mellon university fly ash visual feature boundary detection video retrieval diabetes mellitu semantic indexing oryza sativa water storage user association efficient wireles shot boundary shot boundary detection data assimilation system retrieval task controlled trial terrestrial television video search gps network sensor network consist efficient wireless sensor information retrieval task concept detection video captioning retrieval evaluation rice seed safety equipment endangered species station operation case study involving dublin city university high-level feature seed germination brown coal high plain study involving structure recognition climate experiment gravity recovery table structure land data assimilation instance search combinatorial number randomised controlled trial recovery and climate randomised controlled combinatorial number theory adult male high-level feature extraction complete proof music perception robust computation optimization-based method perception and cognition global land datum social perception terrestrial water storage trec video retrieval terrestrial water object-oriented conceptual video retrieval evaluation trec video seed variety base station operation table structure recognition transgenic rice concept detector total water storage groundwater storage regional gp grace gravity randomized distributed algorithm ibm tivoli workload scheduler cerebrovascular accident case study united state