Detecting semantic concepts from video using temporal gradients and audio classification

Mika Rautiainen, Tapio Seppänen, Jani Penttilä, Johannes Peltola

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingsScientificpeer-review

8 Citations (Scopus)


In this paper we describe new methods to detect semantic concepts from digital video based on audible and visual content. Temporal Gradient Correlogram captures temporal correlations of gradient edge directions from sampled shot frames. Power-related physical features are extracted from short audio samples in video shots. Video shots containing people, cityscape, landscape, speech or instrumental sound are detected with trained self-organized maps and kNN classification results of audio samples. Test runs and evaluations in TREC 2002 Video Track show consistent performance for Temporal Gradient Correlogram and state-of-the-art precision in audio-based instrumental sound detection.
Original languageEnglish
Title of host publicationImage and Video Retrieval
Subtitle of host publicationCIVR 2003
ISBN (Electronic)978-3-540-45113-6
ISBN (Print)978-3-540-40634-1
Publication statusPublished - 2003
MoE publication typeA4 Article in a conference publication
EventImage and Video Retrieval, CIVR 2003 - Urbana-Champaign, United States
Duration: 24 Jul 200325 Jul 2003

Publication series

SeriesLecture Notes in Computer Science


ConferenceImage and Video Retrieval, CIVR 2003
Abbreviated titleCIVR 2003
CountryUnited States

Fingerprint Dive into the research topics of 'Detecting semantic concepts from video using temporal gradients and audio classification'. Together they form a unique fingerprint.

Cite this