• Ms. Bharti Vishwakarma
• Mr. Rajneesh Pachouri
• Mr. Anurag Jain
• Ms. Sadhvi Biltharey
• Mrs Monali Sahoo

The rapid evolution of deep fake technology has posed serious threats to digital authenticity, privacy, and public trust. Deep fake videos, generated using advanced generative models, can convincingly manipulate facial expressions and voice, making it increasingly difficult to distinguish real content from fake. This thesis presents a robust deep fake video detection framework that integrates Convolutional Neural Networks (CNN) with Recurrent Neural Networks (RNN), specifically combining AlexNet for spatial feature extraction and Long Short-Term Memory (LSTM) networks for capturing temporal patterns across video frames. The proposed system involves video frame extraction, face detection, and pre-processing steps to standardize inputs. A custom dataloader feeds these inputs into the hybrid model architecture, improving detection accuracy by leveraging both spatial and temporal dependencies. The model is trained, validated, and tested on a comprehensive dataset, and also supports real-time video uploads for prediction. Experimental results demonstrate significant improvements in precision, recall, and F1-score compared to traditional CNN-only approaches. This approach shows great promise in strengthening automated defenses against deep fake content.

Deepfake Detection, AlexNet, LSTM, Temporal Features, Video Forensics, Face Pre-processing

"A Hybrid CNN-LSTM Framework for Robust Deep Fake Video Detection", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 9, page no.f202-f207, September-2025, Available :http://www.jetir.org/papers/JETIR2509529.pdf

"A Hybrid CNN-LSTM Framework for Robust Deep Fake Video Detection", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 9, page no. ppf202-f207, September-2025, Available at : http://www.jetir.org/papers/JETIR2509529.pdf