Papers
Topics
Authors
Recent
Assistant
AI Research Assistant
Well-researched responses based on relevant abstracts and paper content.
Custom Instructions Pro
Preferences or requirements that you'd like Emergent Mind to consider when generating responses.
Gemini 2.5 Flash
Gemini 2.5 Flash 71 tok/s
Gemini 2.5 Pro 54 tok/s Pro
GPT-5 Medium 24 tok/s Pro
GPT-5 High 25 tok/s Pro
GPT-4o 124 tok/s Pro
Kimi K2 200 tok/s Pro
GPT OSS 120B 463 tok/s Pro
Claude Sonnet 4.5 37 tok/s Pro
2000 character limit reached

A Convolutional LSTM based Residual Network for Deepfake Video Detection (2009.07480v1)

Published 16 Sep 2020 in cs.CV and cs.MM

Abstract: In recent years, deep learning-based video manipulation methods have become widely accessible to masses. With little to no effort, people can easily learn how to generate deepfake videos with only a few victims or target images. This creates a significant social problem for everyone whose photos are publicly available on the Internet, especially on social media websites. Several deep learning-based detection methods have been developed to identify these deepfakes. However, these methods lack generalizability, because they perform well only for a specific type of deepfake method. Therefore, those methods are not transferable to detect other deepfake methods. Also, they do not take advantage of the temporal information of the video. In this paper, we addressed these limitations. We developed a Convolutional LSTM based Residual Network (CLRNet), which takes a sequence of consecutive images as an input from a video to learn the temporal information that helps in detecting unnatural looking artifacts that are present between frames of deepfake videos. We also propose a transfer learning-based approach to generalize different deepfake methods. Through rigorous experimentations using the FaceForensics++ dataset, we showed that our method outperforms five of the previously proposed state-of-the-art deepfake detection methods by better generalizing at detecting different deepfake methods using the same model.

Citations (71)

Summary

We haven't generated a summary for this paper yet.

Lightbulb Streamline Icon: https://streamlinehq.com

Continue Learning

We haven't generated follow-up questions for this paper yet.

List To Do Tasks Checklist Streamline Icon: https://streamlinehq.com

Collections

Sign up for free to add this paper to one or more collections.

Don't miss out on important new AI/ML research

See which papers are being discussed right now on X, Reddit, and more:

“Emergent Mind helps me see which AI papers have caught fire online.”

Philip

Philip

Creator, AI Explained on YouTube