Hello! My name is... Buffy'' -- Automatic Naming of Characters in TV Video

Everingham, Mark; Šivic, Josef; Zisserman, Andrew

doi:10.5244/c.20.92

Cited by 483 publications

(572 citation statements)

References 19 publications

Supporting

Mentioning

560

Contrasting

Unclassified

Order By: Relevance

“…For instance, visual person models could be improved by adding clothes descriptors [20] or by introducing pose information [21] in the face comparison functions, while on the association side, the module could be improved using contextual information given by role recognition [22] or shot classes (studio vs field, closeups, group, public, etc ) automatically derived from scene content descriptors, movements, duration...…”

Section: Resultsmentioning

confidence: 99%

A conditional random field approach for audio-visual people diarization

Khoury

Meignier²,

Odobez

et al. 2014

2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

We investigate the problem of audio-visual (AV) person diarization in broadcast data. That is, automatically associate the faces and voices of people and determine when they appear or speak in the video. The contributions are twofolds. First, we formulate the problem within a novel CRF framework that simultaneously performs the AV association of voices and face clusters to build AV person models, and the joint segmentation of the audio and visual streams using a set of AV cues and their association strength. Secondly, we use for this AV association strength a score that does not only rely on lips activity, but also on contextual visual information (face size, position, number of detected faces,. . . ) that leads to more reliable association measures. Experiments on 6 hours of broadcast data show that our framework is able to improve the AV-person diarization especially for speaker segments erroneously labeled in the mono-modal case.

show abstract

Section: Resultsmentioning

confidence: 99%

A conditional random field approach for audio-visual people diarization

Khoury

Meignier²,

Odobez

et al. 2014

2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

show abstract

“…the global mean matrix k 1 the length of 1D signature k 2 the length of 2D signature smaller than d and n is smaller than m. We call the k 2 × n matrix "2D signature". For easy reference, main symbols used in the paper are listed in Table 1.…”

Section: Discriminative Signature Generationmentioning

confidence: 99%

Efficient Large-Scale Video Retrieval via Discriminative Signatures

Hao

Kamata

2013

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

Pengyi HAO †a) , Student Member and Sei-ichiro KAMATA †b) , Member SUMMARYThe topic of retrieving videos containing a desired person from a dataset just using the content of faces without any help of textual information has many interesting applications like video surveillance, social network, video mining, etc. However, traditional face matching against a huge number of detected faces leads to an unacceptable response time and may also reduce the accuracy due to the large variations in facial expressions, poses, lighting, etc. Therefore, in this paper we propose a novel method to generate discriminative "signatures" for efficiently retrieving the videos containing the same person with a query. In this research, the signature is defined as a compact, discriminative and reduced dimensionality representation, which is generated from a set of high-dimensional feature vectors of an individual. The desired videos are retrieved based on the similarities between the signature of the query and those of individuals in the database. In particular, we make the following contributions. Firstly, we give an algorithm of two directional linear discriminant analysis with maximum correntropy criterion (2DLDA-MCC) as an extension to our recently proposed maximum correntropy criterion based linear discriminant analysis (LDA-MCC). Both algorithms are robust to outliers and noise. Secondly, we present an approach for transferring a set of exemplars to a fixed-length signature using LDA-MCC and 2DLDA-MCC, resulting in two kinds of signatures that are called 1D signature and 2D signature. Finally, a novel video retrieval scheme is given based on the signatures, which has low storage requirement and can achieve a fast search. Evaluations on a large dataset of videos show reliable measurement of similarities by using the proposed signatures to represent the identities generated from videos. Experimental results also demonstrate that the proposed video retrieval scheme has the potential to substantially reduce the response time and slightly increase the mean average precision of retrieval. key words : individual retrieval, signature, linear discriminant analysis, video dataset

show abstract

“…Some of the works are either limited to some specific domains (e.g. movies [12,13], TV videos [14,15,16] etc.) or focus on certain predefined content such as human face [17,18] and human activities [19].…”

Section: Related Workmentioning

confidence: 99%

Tagging based Efficient Web Video Event Categorization

Raghavendra¹

2014

IJCATR

View full text Add to dashboard Cite

Web video categorization is one of the emerging research fields in the computer vision domain due to its massive volume growth in the internet which demands to discover events. Due to insufficient, noisy information and large intra class disparity makes it more daunting task to recognize the events. Most of the recent works focus on constrained (fixed camera, known environment) videos with supervised labelling to categorize the web videos. In this paper, we propose the subject based Part-Of-Speech (POS) Tagging technique with the assist of Named Entity Recognition (NER) and WordNet is applied on YouTube video titles to discover the events based on the subject, not on the objects visualized in the videos. Unsupervised learning method is used on high level features (titles) because of incoming videos are not known and large intra-class variations. For the experiment, we have chosen topics from Google Zeitgeist and downloaded the related videos from YouTube. A novel conclusion is derived from the experimental result that use of low level features will lead to a poor classification in discovering intra class events based on the subject of the videos.

show abstract

Hello! My name is... Buffy'' -- Automatic Naming of Characters in TV Video

Cited by 483 publications

References 19 publications

A conditional random field approach for audio-visual people diarization

A conditional random field approach for audio-visual people diarization

Efficient Large-Scale Video Retrieval via Discriminative Signatures

Tagging based Efficient Web Video Event Categorization

Contact Info

Product

Resources

About