國立高雄大學圖資館 |

登入

回首頁

Enhancing Professional Communication Through AI-Powered Refinement of Video Content /

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Enhancing Professional Communication Through AI-Powered Refinement of Video Content /Yanchi Liu.
作者:	Liu, Yanchi,
面頁冊數:	1 electronic resource (71 pages)
附註:	Source: Masters Abstracts International, Volume: 86-07.
附註:	Advisors: Zhao, Yijun Committee members: Weiss, Gary; Li, Yanjun.
Contained By:	Masters Abstracts International86-07.
標題:	Computer science.
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31768163
ISBN:	9798302828446

Enhancing Professional Communication Through AI-Powered Refinement of Video Content /
Liu, Yanchi,

Enhancing Professional Communication Through AI-Powered Refinement of Video Content /Yanchi Liu. - 1 electronic resource (71 pages)

Source: Masters Abstracts International, Volume: 86-07.

The rising demand for high-quality video content across educational, professional, and social media platforms has highlighted the need for efficient solutions to enhance presentation quality, particularly for non-native speakers. While artificial intelligence has made significant strides in various aspects of content processing, existing video enhancement solutions often require substantial manual intervention and struggle to maintain natural speech patterns while improving clarity. This research presents an innovative automated pipeline, Video GPT, that integrates state-of-the-art AI technologies to refine video content while preserving speaker authenticity and ensuring precise Audio-Video Synchronization.Our system implements a dual-track approach to address the distinct challenges presented by different video types. For presenter-free content, such as slide presentations, the base pipeline focuses on aligning refined audio with visual transitions and animations. For presenter-present videos, where lip synchronization and facial expressions are critical, we developed an enhanced pipeline featuring advanced segmentation strategies and sophisticated alignment algorithms. Both tracks leverage automatic speech recognition for precise timestamping, large language models for text refinement, and voice cloning technology to maintain speaker identity while improving clarity.Video GPT processes a three-minute video in under 17 minutes, making it practical for real-world applications. The system effectively normalizes speech rates, corrects grammatical errors, eliminates filler words, and reduces accent-related issues while preserving the speaker's unique vocal characteristics. These capabilities make it particularly valuable for educational content creation, non-native speaker support, and professional media production, offering a powerful tool for improving content accessibility and engagement across global audiences.

English

ISBN: 9798302828446Subjects--Topical Terms:

199325
Computer science.
Subjects--Index Terms:

Automatic Speech Recognition

Enhancing Professional Communication Through AI-Powered Refinement of Video Content /
LDR:03396nmm a22004453i 4500 001 676415
005 20250420214402.5
006 m o d
007 cr|nu||||||||
008 250901s2024 miu||||||m |||||||eng d
020 $a 9798302828446
035 $a (MiAaPQD)AAI31768163
035 $a AAI31768163
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Liu, Yanchi, $e author. $3 990251
245 1 0 $a Enhancing Professional Communication Through AI-Powered Refinement of Video Content / $c Yanchi Liu.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2024
300 $a 1 electronic resource (71 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Masters Abstracts International, Volume: 86-07.
500 $a Advisors: Zhao, Yijun Committee members: Weiss, Gary; Li, Yanjun.
502 $b M.S. $c Fordham University $d 2024.
520 $a The rising demand for high-quality video content across educational, professional, and social media platforms has highlighted the need for efficient solutions to enhance presentation quality, particularly for non-native speakers. While artificial intelligence has made significant strides in various aspects of content processing, existing video enhancement solutions often require substantial manual intervention and struggle to maintain natural speech patterns while improving clarity. This research presents an innovative automated pipeline, Video GPT, that integrates state-of-the-art AI technologies to refine video content while preserving speaker authenticity and ensuring precise Audio-Video Synchronization.Our system implements a dual-track approach to address the distinct challenges presented by different video types. For presenter-free content, such as slide presentations, the base pipeline focuses on aligning refined audio with visual transitions and animations. For presenter-present videos, where lip synchronization and facial expressions are critical, we developed an enhanced pipeline featuring advanced segmentation strategies and sophisticated alignment algorithms. Both tracks leverage automatic speech recognition for precise timestamping, large language models for text refinement, and voice cloning technology to maintain speaker identity while improving clarity.Video GPT processes a three-minute video in under 17 minutes, making it practical for real-world applications. The system effectively normalizes speech rates, corrects grammatical errors, eliminates filler words, and reduces accent-related issues while preserving the speaker's unique vocal characteristics. These capabilities make it particularly valuable for educational content creation, non-native speaker support, and professional media production, offering a powerful tool for improving content accessibility and engagement across global audiences.
546 $a English
590 $a School code: 0072
650 4 $a Computer science. $3 199325
650 4 $a Information technology. $3 184390
653 $a Automatic Speech Recognition
653 $a Large Language Models
653 $a Pipeline
653 $a Refinement
653 $a Synchronization
653 $a Text-to-Speech
690 $a 0984
690 $a 0489
690 $a 0800
710 2 $a Fordham University. $b Computer and Information Science. $3 708593
720 1 $a Zhao, Yijun $e degree supervisor.
773 0 $t Masters Abstracts International $g 86-07.
790 $a 0072
791 $a M.S.
792 $a 2024
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31768163