Abstract: An image capture device captures video frames using a capture rate. The image capture device performs face detection on the video frames using a face detection rate slower than the capture rate. Temporal filtering is applied to determine the positions and sizes of the faces within the video frames in which face detection has not been performed. Tracking filtering is applied to determine the positions and sizes of the faces within the video frames in which face detection has failed. One or more of the faces within the video frames are blurred, and the blurred video frames are transmitted to a computing device.

Abstract: Systems, apparatus, and methods for dynamic encoder configuration. In one exemplary embodiment, a machine-learning model uses pixel features and encoding features from previous stages of an image processing pipeline (IPP) to dynamically adjust bitrate. The machine-learning model is trained to select bitrate adjustments for an encoder such that the expected image quality of a video stream remains at a selected quality level (e.g., SSIM, VMAF, VIF, HVS-PSNR, etc.). Conventional dynamic encoding solutions are focused on encode-once-deliver-often (best-effort) applications, the exemplary IPP is designed for real-time applications that may not have the benefit of actual subsequent encoding quality analysis; instead proxy data (pixel features and encoding features) that are representative approximations of image complexity are used.