Amplifying Video Information-Seeking Success through Rich, Exploratory Interfaces

Years of international participation and collaboration in TRECVID have shown that interactive multimedia systems, those with a user in the search loop, have consistently outperformed fully automated systems. Interface capabilities like querying by text, by image, and by semantic concept and storyboard layouts have led to significant performance improvements on provided search tasks. Lessons learned for TRECVID shot-based video retrieval are presented. In the real world, however, video collection users may focus on story threads instead of shots, or may not be provided with a clear stated search task. The paper also discusses users facing situations where they lack the knowledge or contextual awareness to formulate queries and having a genuine need for exploratory search systems supporting serendipitous browsing. Interfaces promoting various views for navigating complex information spaces can help with exploratory search and investigation into video corpora ranging from documentaries to broadcast news to oral histories. Work is presented using The HistoryMakers oral history archive as well as TRECVID international broadcast news to discuss the utility of various query and presentation mechanisms emphasizing people, time, location, and visual attributes. The paper leads into a discussion of how exploratory interfaces for video extend beyond storyboards, with a series of user studies referenced as empirical data in support of the presented conclusions.

[1]  Michael G. Christel Evaluation and user studies with respect to video summarization and browsing , 2006, Electronic Imaging.

[2]  Marcel Worring,et al.  Mediamill: Advanced Browsing in News Video Archives , 2006, CIVR.

[3]  Peter G. B. Enser,et al.  Retrieval of Archival Moving Imagery - CBIR Outside the Frame? , 2002, CIVR.

[4]  Ramesh C. Jain,et al.  ACM SIGMM retreat report on future directions in multimedia research , 2005, TOMCCAP.

[5]  Michael G. Christel,et al.  Improving Access to a Digital Video Library , 1997, INTERACT.

[6]  Michael G. Christel Establishing the utility of non-text search for news video retrieval with real world users , 2007, ACM Multimedia.

[7]  Zygmunt Pizlo,et al.  Automated video program summarization using speech transcripts , 2006, IEEE Transactions on Multimedia.

[8]  Michael G. Christel,et al.  Mining Novice User Activity with TRECVID Interactive Retrieval Tasks , 2006, CIVR.

[9]  Michael G. Christel,et al.  Addressing the challenge of visual information access from digital image and video libraries , 2005, Proceedings of the 5th ACM/IEEE-CS Joint Conference on Digital Libraries (JCDL '05).

[10]  Wei-Ying Ma,et al.  Image and Video Retrieval , 2003, Lecture Notes in Computer Science.

[11]  Ben Shneiderman,et al.  Strategies for evaluating information visualization tools: multi-dimensional in-depth long-term case studies , 2006, BELIV '06.

[12]  Michael G. Christel,et al.  Finding the right shots: assessing usability and performance of a digital video library interface , 2004, MULTIMEDIA '04.

[13]  Michael G. Christel,et al.  Evaluating the contributions of video representation for a life oral history collection , 2008, JCDL '08.

[14]  Wolfgang Effelsberg,et al.  Video abstracting , 1997, CACM.

[15]  Marcel Worring,et al.  A Learned Lexicon-Driven Paradigm for Interactive Video Retrieval , 2007, IEEE Transactions on Multimedia.

[16]  Rong Yan,et al.  Merging storyboard strategies and automatic retrieval for improving interactive video search , 2007, CIVR '07.

[17]  Alexander G. Hauptmann,et al.  Successful approaches in the TREC video retrieval evaluations , 2004, MULTIMEDIA '04.