A Cross-Modal Neural Architecture for Real-Time Deepfake Detection to Mitigate Misinformation

Deepfake content is spreading quickly online, making it hard to know what’s real. Because of this, we need strong tools that can detect fake images, audio, and videos. In this study, we created a three-step deep learning system that can check all three types of data. In the first step, we used an EfficientNet model to detect fake images. This model is good at noticing tiny details that reveal if an image is fake. After training it with a Kaggle dataset, it reached an accuracy of 95.56%. The second step focuses on fake audio. We have used CNN model to build it which helps the system understand patterns in sound and related facial features. This model was trained on the DEEP-VOICE dataset and achieved 98.5% accuracy.The third step looks at fake videos. In this model we have combinedEfficientNet with an LSTM and a Random Forest classifier. This model helps in both audio and video frames to reach the accuracy of 95.56%. Finally, we combined all three models into one application that can check images, audio, and video in real time. Overall shows that by using multimodel it helps the deep fake to detect the in all the forms.

Authors

Institutions

Publication Details

Journal
The International Journal of Maritime Engineering
Published
2026-11-13
DOI
https://doi.org/10.5750/ijme.v168ia6(s).2019
Primary Topic
Digital Media Forensic Detection
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

A Cross-Modal Neural Architecture for Real-Time Deepfake Detection to Mitigate Misinformation

Mallikarjun Gachchannavar, J. R. Naveen Kumar
2 links
The International Journal of Maritime Engineering
Digital Media Forensic Detection
article

A Cross-Modal Neural Architecture for Real-Time Deepfake Detection to Mitigate Misinformation

Mallikarjun Gachchannavar, J. R. Naveen Kumar
article en

Abstract

Deepfake content is spreading quickly online, making it hard to know what’s real. Because of this, we need strong tools that can detect fake images, audio, and videos. In this study, we created a three-step deep learning system that can check all three types of data. In the first step, we used an EfficientNet model to detect fake images. This model is good at noticing tiny details that reveal if an image is fake. After training it with a Kaggle dataset, it reached an accuracy of 95.56%. The second step focuses on fake audio. We have used CNN model to build it which helps the system understand patterns in sound and related facial features. This model was trained on the DEEP-VOICE dataset and achieved 98.5% accuracy.The third step looks at fake videos. In this model we have combinedEfficientNet with an LSTM and a Random Forest classifier. This model helps in both audio and video frames to reach the accuracy of 95.56%. Finally, we combined all three models into one application that can check images, audio, and video in real time. Overall shows that by using multimodel it helps the deep fake to detect the in all the forms.

The International Journal of Maritime EngineeringVol. 168(A6(S))
Srinivas University (IN)
Digital Media Forensic Detection
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon