AI-powered face and body blur for Zoom, Teams, Google Meet, OBS, and any app that uses a webcam. Virtual camera output — no plugins needed. GPU-accelerated, on-premise, real-time.
SmartBlur runs AI detection on every frame of your live camera feed with near-zero latency. Faces, heads, and bodies are detected and blurred as they appear on screen — no post-processing, no delay. The detection engine processes video at full framerate so there are no dropped frames or visible lag between motion and blur.
SmartBlur creates a virtual camera on your system that any application can use. Open Zoom, Microsoft Teams, Google Meet, OBS Studio, or any video conferencing tool, and select "SmartBlur Virtual Camera" as your video source. Every participant sees the blurred feed — you see your original camera unmodified on your own screen.
Connect any USB webcam, built-in laptop camera, or RTSP/IP network camera stream. SmartBlur accepts standard camera inputs and processes them in real-time. Use a high-end PTZ security camera, a budget USB webcam, or an IP camera on your network — all are supported out of the box with no extra configuration.
NVIDIA GPU acceleration keeps the blur pipeline fast enough for live use. The AI model runs inference on your GPU while the CPU handles camera I/O and virtual camera output — a parallel pipeline designed for sustained real-time performance. Falls back to CPU on systems without a compatible GPU.
Change blur type, intensity, detection target, and confidence thresholds while the camera is running. Switch from Gaussian to Pixelated blur mid-stream, increase padding to cover more of the head, or toggle between face-only and full-body detection — all changes apply instantly without restarting the camera.
Your live camera feed never leaves your machine. SmartBlur processes everything locally — no frames are sent to any server, no internet connection is required during use. The virtual camera output is generated entirely on your hardware. This makes SmartBlur suitable for confidential environments where cloud-based blurring tools are not permitted.
Fine-tune detection and blur in real-time. All changes apply instantly without restarting the camera.
Gaussian for smooth blur, Pixelated for mosaic, Solid fill for blackout, or Overlay to stamp a custom image. Change mid-stream without restarting.
Control how strong the blur appears. Low for a soft haze, high for complete unrecognizability. Adjust in real-time.
Face (360-degree coverage including sides and back of head) or full body. Each target has its own confidence threshold.
Separate sliders for face and body. Lower catches more detections in difficult lighting, higher reduces false positives.
Expand the blur region beyond the detected bounding box (0–50%). Ensures ears, hair, and chins are fully covered in every frame.
Pick your input source — USB webcam, built-in camera, or enter an RTSP/IP camera URL. SmartBlur auto-detects available devices.
Open SmartBlur and select your USB webcam, built-in camera, or enter an RTSP stream URL. The camera feed appears in the SmartBlur window instantly.
Choose what to detect (face or body), set the confidence threshold, pick a blur style, and adjust intensity. You see the blurred preview in the SmartBlur window in real-time.
Click Start to activate the virtual camera output. SmartBlur creates a system-level virtual webcam device that mirrors your blurred feed. Your original camera remains untouched.
Open Zoom, Teams, OBS, or any video app. Go to camera settings and choose "SmartBlur Virtual Camera" as your video source. That's it — everyone on the call sees the blurred feed while you see the original.
SmartBlur creates a standard virtual camera device. If your app can see a webcam, it can use SmartBlur.
Blur your face or background participants during Zoom, Teams, and Google Meet calls. Protect family members who walk through the frame, or maintain personal privacy during internal meetings without disabling your camera entirely.
Stream on YouTube, Twitch, or Facebook Live through OBS Studio with automatic face blur on the virtual camera. Protect bystanders, minors, or confidential locations in outdoor IRL streams without manual masking.
Healthcare providers can anonymize patient faces during telehealth sessions for training, supervision, or research recording. Patients can blur their own face during group therapy to maintain anonymity while still participating visually.
Educators can blur student faces during recorded virtual classrooms for FERPA compliance. Guest speakers and panelists can maintain visual anonymity while still presenting on camera.
Connect RTSP security camera feeds to SmartBlur and output an anonymized virtual camera feed. Operators can monitor activity without exposing individual identities, supporting privacy-by-design principles in surveillance operations.
Live-stream interviews and on-the-ground reporting with automatic face blur to protect sources, whistleblowers, and bystanders. The virtual camera works with any broadcasting app that accepts a camera input.
Free trial available. One-time payment. No subscriptions.
Get SmartBlur