Character-aware audio-visual subtitling in context
This paper presents an improved framework for character-aware audio-visual subtitling in TV shows. Our approach integrates speech recognition, speaker diarisation, and character recognition, utilising both audio and visual cues. This holistic solution addresses what is said, when it’s said, and who...
| Main Authors: | , |
|---|---|
| Format: | Conference item |
| Language: | English |
| Published: |
Springer
2024
|
| _version_ | 1824458577886052352 |
|---|---|
| author | Huh, J Zisserman, A |
| author_facet | Huh, J Zisserman, A |
| author_sort | Huh, J |
| collection | OXFORD |
| description | This paper presents an improved framework for character-aware audio-visual subtitling in TV shows. Our approach integrates speech recognition, speaker diarisation, and character recognition, utilising both audio and visual cues. This holistic solution addresses what is said, when it’s said, and who is speaking, providing a more comprehensive and accurate character-aware subtitling for TV shows. Our approach brings improvements on two fronts: first, we show that audio-visual synchronisation can be used to pick out the talking face amongst others present in a video clip, and assign an identity to the corresponding speech segment. This audio-visual approach improves recognition accuracy and yield over current methods. Second, we show that the speaker of short segments can be determined by using the temporal context of the dialogue within a scene. We propose an approach using local voice embeddings of the audio, and large language model reasoning on the text transcription. This overcomes a limitation of existing methods that they are unable to accurately assign speakers to short temporal segments. We validate the method on a dataset with 12 TV shows, demonstrating superior performance in speaker diarisation and character recognition accuracy compared to existing approaches. Project page : https://www.robots.ox.ac.uk/~vgg/research/llr-context/. |
| first_indexed | 2024-12-09T03:16:02Z |
| format | Conference item |
| id | oxford-uuid:439eec3e-fc01-4f09-b401-b03d88b931b5 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2025-02-19T04:28:07Z |
| publishDate | 2024 |
| publisher | Springer |
| record_format | dspace |
| spelling | oxford-uuid:439eec3e-fc01-4f09-b401-b03d88b931b52024-12-13T09:37:38ZCharacter-aware audio-visual subtitling in contextConference itemhttp://purl.org/coar/resource_type/c_5794uuid:439eec3e-fc01-4f09-b401-b03d88b931b5EnglishSymplectic ElementsSpringer2024Huh, JZisserman, AThis paper presents an improved framework for character-aware audio-visual subtitling in TV shows. Our approach integrates speech recognition, speaker diarisation, and character recognition, utilising both audio and visual cues. This holistic solution addresses what is said, when it’s said, and who is speaking, providing a more comprehensive and accurate character-aware subtitling for TV shows. Our approach brings improvements on two fronts: first, we show that audio-visual synchronisation can be used to pick out the talking face amongst others present in a video clip, and assign an identity to the corresponding speech segment. This audio-visual approach improves recognition accuracy and yield over current methods. Second, we show that the speaker of short segments can be determined by using the temporal context of the dialogue within a scene. We propose an approach using local voice embeddings of the audio, and large language model reasoning on the text transcription. This overcomes a limitation of existing methods that they are unable to accurately assign speakers to short temporal segments. We validate the method on a dataset with 12 TV shows, demonstrating superior performance in speaker diarisation and character recognition accuracy compared to existing approaches. Project page : https://www.robots.ox.ac.uk/~vgg/research/llr-context/. |
| spellingShingle | Huh, J Zisserman, A Character-aware audio-visual subtitling in context |
| title | Character-aware audio-visual subtitling in context |
| title_full | Character-aware audio-visual subtitling in context |
| title_fullStr | Character-aware audio-visual subtitling in context |
| title_full_unstemmed | Character-aware audio-visual subtitling in context |
| title_short | Character-aware audio-visual subtitling in context |
| title_sort | character aware audio visual subtitling in context |
| work_keys_str_mv | AT huhj characterawareaudiovisualsubtitlingincontext AT zissermana characterawareaudiovisualsubtitlingincontext |