Authors: Lana Touma, Mohammad Al Horani, Manar Tailouni, Anas Dahabiah, Khloud Al Jallad
Published on: June 30, 2023
Impact Score: 7.2
Arxiv code: Arxiv:2307.07516
Summary
- What is new: A new voting-based method integrating audio, visual, and lexical features for deception detection.
- Why this is important: Automatic deception detection from videos has always been challenging.
- What the research proposes: A multimodal approach that uses CNN, SVM, and Word2Vec on different modalities (images, audio, manuscripts) for deception detection.
- Results: Achieved high accuracy in deception detection: 97%, 96%, 92% on the Real-Life Trial Dataset and 97%, 82%, 73% on the Miami University Deception Detection dataset.
Technical Details
Technological frameworks used: nan
Models used: CNN, SVM, Word2Vec
Data used: Real-life trial dataset by Michigan University, Miami University deception detection dataset
Potential Impact
Security firms, law enforcement, human resources, and online platforms could benefit or be disrupted.
Want to implement this idea in a business?
We have generated a startup concept here: TruthLens.
Leave a Reply