Real-event-anchored benchmark for detecting AI-generated videos in real-world crisis settings.
-
Updated
Aug 17, 2026 - Python
Real-event-anchored benchmark for detecting AI-generated videos in real-world crisis settings.
EfficientNet+MTCNN deepfake detection achieving 87% accuracy with 12.96% EER - Multi-modal video forensics for misinformation prevention
🛡️ Multi-layer KYC video integrity system for detecting deepfakes, replay attacks, face manipulation, and presentation attacks with explainable risk scoring.
A deepfake detection platform using fine-tuned EfficientNet-B4. Features GradCAM explainability to show exactly where a video was manipulated, alongside an InsightFace pipeline.
Next-gen video platform.
HLS Stream Analyzer – A passive, security-focused tool to analyze and reconstruct HLS media streams, helping researchers identify exposed segments, misconfigurations, and potential risks in streaming infrastructure.
A tool that turns live cameras into a real-time threat-detection and search command center
Production-grade deepfake detection — 3-model ensemble (Xception + EfficientNet-B4 + ViT), calibrated confidence, forensics suite, REST API, and video AV-sync/identity tracking.
fractalVideoGuard
aI-powered multimedia forensic analysis system for detecting CCTV tampering, deepfakes, and metadata manipulation.
Lip-sync deepfake detection built on LipForensics, with LLM (Gemini) guided video segment selection and documented preprocessing experiments.
Frequency-aware deepfake detection under lossy H.264 compression using XceptionNet, DCT features, learnable frequency masking, and cross-attention fusion.
Forensic-first DVR/NVR storage reconstruction, triage and CCTV video recovery platform
Video forensics for deepfakes, AI-generated content, and miscontextualized footage
A client-side, air-gapped web application for digital forensic examiners to calculate CCTV temporal variance and synchronize video timelines.
Cryptographic proof of video authenticity. Verify video origin and integrity using perceptual hashing and digital signatures to combat manipulated media.
Forensic detection engine for AI-generated video. Dual-path ensemble over noise-residual, DCT frequency, temporal, and CLIP signals, with GradCAM explainability and LLM-driven forensic reports.
Training-free deepfake and face-swap detection. Explainable forensic analysis of images and video — no model download, no GPU, runs offline. The inverse of deepfakes/faceswap.
Experimental 8-layer heuristic detector for AI-generated video: frequency, face symmetry, motion, edges, texture, lighting, audio sync, metadata
SaksiAI is an enterprise-grade Evidence Intelligence Platform (EIP) specifically designed for the Indonesian market. It addresses the Hidden Data Problem where critical organizational knowledge is trapped in unusable, fragmented formats such as scanned PDFs, voice recordings, and CCTV feeds.
To associate your repository with the video-forensics topic, visit your repo's landing page and select "manage topics."