Linking identities and viewpoints in home movies based on robust feature matching

The identification of useful structures in home video is difficult because this class of video is distinguished from other video sources by its unrestricted, non edited content and the absence of regulated storyline. In addition, home videos contain a lot of motion and erratic camera movements, with shots of the same character being captured from various angles and viewpoints. In this paper, we present a solution to the challenging problem of clustering shots and faces in home videos, based on the use of SIFT features. SIFT features have been known to be robust for object recognition; however, in dealing with the complexities of home video setting, the matching process needs to be augmented and adapted. This paper describes various techniques that can improve the number of matches returned as well as the correctness of matches. For example, existing methods for verification of matches are inadequate for cases when a small number of matches are returned, a common situation in home videos. We address this by constructing a robust classifier that works on matching sets instead of individual matches, allowing the exploitation of the geometric constraints between matches. Finally, we propose techniques for robustly extracting target clusters from individual feature matches.

Witness recall across repeated interviews in a case of repeated abuse.

In this illustrative case study we examine the three forensic interviews of a girl who experienced repeated sexual abuse from ages 7 to 11. She disclosed the abuse after watching a serialized television show that contained a storyline similar to her own experience. This triggered an investigation that ended in successful prosecution of the offender. Because this case involved abuse that was repeated on a weekly basis for 4 years we thus investigated the degree to which the child's narrative reflected specific episodes or generic accounts, and both the interviewer's and child's attempts to elicit and provide, respectively, specific details across the 3 interviews collected in a 1 month period. Across the 3 interviews, the child's account was largely generic, yet on a number of occasions she provided details specific to individual incidents (episodic leads) that could have been probed further. As predicted: earlier interviews were characterized more by episodic than generic prompts and the reverse was true for the third interview; the child often responded using the same style of language (episodic or generic) as the interviewer; and open questions yielded narrative information. We discuss the importance of adopting children's words to specify occurrences, and the potential benefits of permitting generic recall in investigative interviews on children's ability to provide episodic leads. Despite the fact that the testimony was characterized by generic information about what usually happened, rather than specific episodic details about individual occurrences, this case resulted in successful prosecution.