Portfolio case study
Face Antispoofing & Multi-Modal Vision-Language Models
CLIP-driven anti-spoofing experiments for secure face auth.
Overview
Explores face anti-spoofing with multi-modal encoders (text + image cues) to flag replays/deepfakes in authentication flows.
What It Covers
- Anti-spoofing
- Multi-modal learning
- Security application
Stack And Topics
- CLIP
- Vision-Language Models
- Biometric Security
Public Signals
- Public references and qualitative delivery signals are used where numeric benchmarks are not public.