• Rutuja Deshmukh
• Dr. Lalit.V.Patil

Multi-Modal summarization is a valuable research tool for both professional and academicians. In this paper, we propose a Multi-Modal summarization method that can automatically generate a textual summary from asynchronous combination of Multi-Modal data such as text, image, audio and video. Our basic goal is to bridge the semantic gap between Multi-Modal content. The summary generation process for textual data includes techniques such as Stemming, Spell correction and N-gram generation. Similarly, for audio data techniques used are Speech transcriptions, which are then converted to text. In the image processing mechanism, the text contained in the images is extracted using OCR (Optical Character Recognition) technique, followed by techniques such as text image matching, topic modeling and theme identification. Similarly, for video processing the video is converted into frames, after which image processing and audio processing techniques are applied to it. Finally, all the Multi-Modal data is collected to generate a textual summary using the LDA (Latent Dirichlet Algorithm). The contribution work is theme identification amongst the visual data. The experimental results show that the Multi-Modal Summarization method outperforms other competitive techniques.

Summarization, Multimedia, Multi-Modal, Cross-Modal, Natural Language Processing, Computer Vision, OCR Technique.

"Automatic Textual Summary Generation Using Multi-Modal Summarization", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.6, Issue 6, page no.559-565, June 2019, Available :http://www.jetir.org/papers/JETIR1906M84.pdf

"Automatic Textual Summary Generation Using Multi-Modal Summarization", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.6, Issue 6, page no. pp559-565, June 2019, Available at : http://www.jetir.org/papers/JETIR1906M84.pdf