Abstract: Multiple Object Tracking (MOT) procedures are used to analyze a video stream to identify and track objects and events of interest across frames in the video stream. According to various embodiments, two or more different models may be separately applied to track an object across multiple video frames. A model may be dynamically evaluated for a frame or group of frames by determining a performance metric for the model, for instance on the level of a frame or group of frames. Then, two or more models may be fused together using a weighting scheme based at least in part on performance metrics for the different models. The fused model may be used to track objects across the frames.

Abstract: One or more objects in a video stream may be selectively in-painted. In-painting refers to the replacement of a portion of a frame or frames in a video stream with updated image data. In-painting may help to protect privacy by replacing an image of a person, document, password, or other sensitive imagery. Multiple object tracking may be used to track objects across different image frames as well as to determine and persist object identity information across the different image frames. In addition, a video stream may be analyzed to identify activities being performed in the video stream. Then, objects may be in-painted depending on factors such as the identity of the object (e.g., a particular person) and/or the activity or activities being performed. For example, a known individual performing a permitted activity may be in-painted, while an unknown individual performing a prohibited activity may not be in-painted.

Abstract: A hardware camera may include a camera sensor configured to determine input image data. The hardware camera may also include an image signal processor configured to perform one or more image signal processing operations on the input image data. The hardware camera may also include a neural processing unit configured to determine encoded image data by encoding the input image data with an image data encoder portion of a camera autoencoder. The camera autoencoder may be trained based on training image data collected from the camera sensor and a fingerprint specific to the hardware camera. The hardware camera may also include a camera communication interface configured to transmit the encoded image data to a remote computing system, which may determine decoded image data by decoding the encoded image data via an image data decoder portion of the camera autoencoder.