Computer Vision for OSINT Research
VFRAME researches and develops state-of-the-art computer vision technologies for application to human rights research and conflict zone monitoring.
Dec 1, 2021: ModelZoo launches. First object detection model preview available for RBK-250 detector, trained using 3D-rendered synthetic data.
Oct 1, 2021: After 3 years of research, VFRAME reaches 99% detection accuracy on the RBK-250 munition using 100% synthetic training data. Case study, white paper, and model to be published in Q4 2021.
Human rights researchers often rely on videos shared online to document war crimes, atrocities, and human rights violations. Manually reviewing these videos is expensive, does not scale, and can cause vicarious trauma. As an increasing number of videos are posted, a new approach is needed to understand these large datasets.
VFRAME has been working with Mnemonig.org, a Berlin-based organization dedicated to documenting war crimes and human rights violations, to develop computer vision tools to address these challenges.
Specifically, VFRAME has been developing a scalable computer vision system for large-scale analysis of video archives to detect illegal munitions, and redaction tools to blur faces. VFRAME can detect objects, such as the RBK-250 munition found in videos from the Syrian conflict, with up to 99% accuracy at over 400 FPS on a single desktop.
Read more about who we are and what we do.