Related Experiment Video
Updated: Oct 3, 2025
09:47
Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches
Published on: December 15, 2023
1.3K
A new design of multimedia big data retrieval enabled by deep feature learning and Adaptive Semantic Similarity
D Sujatha1, M Subramaniam2, Chinnanadar Ramachandran Rene Robin3
1Information Technology, St. Peter's College of Engineering and Technology, Chennai, India.
Multimedia Systems
|February 14, 2022
Summary
This study introduces an Adaptive Semantic Similarity Function (A-SSF) for efficient multimedia big data retrieval. The novel approach enhances deep multimodal hashing by improving semantic feature extraction and selection for better data discovery.
Area of Science:
- Computer Science
- Data Science
- Artificial Intelligence
Background:
- Multimedia big data is rapidly expanding across various sectors, necessitating efficient retrieval methods.
- Current deep multimodal hashing techniques struggle with complex multilevel semantic structures.
- Existing retrieval approaches lack sophisticated methods for handling diverse multimedia data formats and semantic overlaps.
Purpose of the Study:
- To develop an enhanced deep multimedia big data retrieval system using an Adaptive Semantic Similarity Function (A-SSF).
- To address the limitations of existing deep multimodal hashing methods in exploring complex semantic structures.
- To improve the accuracy and efficiency of retrieving relevant multimedia data from large-scale datasets.
Main Methods:
- A hybrid optimization algorithm, Spider Monkey-Deer Hunting Optimization Algorithm (SM-DHOA), was used for semantic feature selection.
- Deep semantic feature extraction was performed using Convolutional Neural Networks (CNN).
- The Adaptive Semantic Similarity Function (A-SSF) was developed to compute correlations between multimedia semantics for retrieval.
Main Results:
- The proposed method demonstrated superior performance across benchmark multimodal datasets.
- The system effectively handles big data distribution using the map-reduce framework in Hadoop.
- Experiments confirmed the proposed method's ability to outperform state-of-the-art retrieval techniques.
Conclusions:
- The Adaptive Semantic Similarity Function (A-SSF) significantly enhances deep multimedia big data retrieval.
- The integration of CNN and SM-DHOA provides robust semantic feature extraction and selection.
- This research offers a promising solution for efficient and accurate large-scale multimedia data discovery.
