V3C1 Dataset: An Evaluation of Content Characteristics

Published

June 5, 2019

Author(s)

George Awad, Fabian Berns, Luca Rossetto, Klaus Sch?ffmann, Christian Beecks

Abstract

In this work we analyze content statistics of the V3C1 dataset, which is the first partition of the Vimeo Creative Commons Collection (V3C). The dataset has been designed to represent true web videos in the wild, with good visual quality and diverse content characteristics, and will serve as evaluation basis for the Video Browser Showdown 2019-2021 and TRECVID Ad-Hoc Video Search tasks 2019-2021. The dataset comes with a shot segmentation (around 1 million shots) for which we analyze content specifics and statistics. %Such an evaluation is relevant for participants of TRECVID AVS or the VBS, as well as for colleagues using this dataset for any other research purpose. Our research will show that the content of V3C1 is very diverse, has no predominant characteristics and provides a low self-similarity. Thus it is very well suited for video retrieval evaluations as well as for participants of TRECVID AVS or the VBS.

Proceedings Title

ACM International Conference on Multimedia Retrieval (ICMR)

Citation

<i>ACM International Conference on Multimedia Retrieval (ICMR)</i> (, Ottawa, , CA, 10-JUN-19 to 13-JUN-19)

Conference Dates

2019-06-10T00:00:00

Conference Location

Ottawa, , CA

Pub Type

Conferences

Download Paper

Local Download

Citation

Awad, G. , Berns, F. , Rossetto, L. , Sch?ffmann, K. and Beecks, C. (2019), V3C1 Dataset: An Evaluation of Content Characteristics, ACM International Conference on Multimedia Retrieval (ICMR), Ottawa, , CA, [online], https://nps-ext.nist.gov/nps/nps_public_api/api/Publication/file/927585 (Accessed November 11, 2026)

Additional citation formats