GARNN-AE-LSTM: A Multimodal Deep Learning Approach for High-Accuracy Video Summarization
Journal of Visualized Experiments · 2025
DOI: 10.3791/69097 ↗Kurz gefasst
Video/Audio in der Zeitschrift Journal of Visualized Experiments, 2025. Verfasst von Jiasheng Jin und 1 weiteren Autor:innen. Beteiligte Einrichtung: Dr. Television School, Sichuan Film and Television University; jiashengjin67@gmail.com. Hauptthemen laut MEDLINE: Deep Learning, Video Recording, Neural Networks, Computer, Image Processing, Computer-Assisted. Die Arbeit nennt 29 Literaturangaben.
Abstract (Originalsprache)
Video summarization focuses on creating concise versions of lengthy videos by preserving essential content. This study reveals a multimodal machine learning strategy that integrates visual and auditory information by using pretrained Gated Recurrent Neural Network architectures referred to as GARNN, combining Gated Recurrent Units (GRUs) and AlexNet, to extract audio, image, and visual features. The keyframe detection is improved by removing redundant frames and applying motion-compensated feature reduction, followed by optional PCA-based dimensionality reduction. An adversarial encoder-based Long Short-Term Memory (AE-LSTM) classifier is employed for temporal modeling by achieving high accuracy in summarization. The results got evaluated by using sensitivity, F-scores, and positive predictive values, and the method attained an average F-score of 0.985. A gated AlexNet is introduced in a multimodal GARNN-AE-LSTM framework, where motion-compensated PCA-based reduction eliminates redundancy, GRUs record temporal progression, and gating fine-tunes spatial feature selection, all of which contribute to a more accurate and efficient video summarization system. The improved F1 score demonstrates the model's effectiveness in generating accuracy in video summaries by creating a meaningful video. This approach highlights the potential of multimodal feature extraction and advanced deep learning techniques for robust video analysis and compression.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Dr. Television School, Sichuan Film and Television University; jiashengjin67@gmail.com.
- Faculty of Creative Arts, University of Malaya.
Zitieren
APA
Jin, J., & Sharudin, S. A. (2025). GARNN-AE-LSTM: A Multimodal Deep Learning Approach for High-Accuracy Video Summarization. Journal of Visualized Experiments. https://doi.org/10.3791/69097
Vancouver
Jin J, Sharudin SA. GARNN-AE-LSTM: A Multimodal Deep Learning Approach for High-Accuracy Video Summarization. Journal of Visualized Experiments. 2025. doi:10.3791/69097
BibTeX
@article{Jin2025,
author = {Jin, Jiasheng and Sharudin, Sharul Azim},
title = {GARNN-AE-LSTM: A Multimodal Deep Learning Approach for High-Accuracy Video Summarization},
journal = {Journal of Visualized Experiments},
year = {2025},
doi = {10.3791/69097}
}
EUVIMED