MuDD: A Multimodal Deception Detection Dataset and GSR-Guided Progressive Distillation for Non-Contact Deception Detection

1. 1Collect and preprocess multimodal data including GSR, visual, and auditory cues
2. 2Develop a cross-modal knowledge distillation framework to guide representation learning in non-contact modalities
3. 3Train and evaluate the model using the MuDD dataset and GSR-Guided Progressive Distillation
4. 4Fine-tune the model for improved performance on non-contact deception detection tasks

AI engineers and researchers on a team can benefit from this work to develop more accurate deception detection models, while data scientists can utilize the MuDD dataset for training and testing

💡 GSR provides reliable physiological cues for deception detection, which can be leveraged to improve non-contact modalities

🔍 Deception detection gets a boost with MuDD dataset & GSR-Guided Progressive Distillation!