Systems, devices, and methods related to a Deep Learning Accelerator and memory are described. For example, a removable media (e.g., a memory card, or a USB drive) may be configured to execute instructions with matrix operands and configured with: an interface to receive a video stream; and random access memory to buffer a portion of the video stream as an input to an Artificial Neural Network and to store instructions executable by the Deep Learning Accelerator and matrices of the Artificial Neural Network. Such a removable media can be used to replace an existing removable media used in a surveillance camera to record video or images. The Deep Learning Accelerator can execute the instructions to generate analytics of the buffer portion using the Artificial Neural Network, enabling the surveillance camera that is upgraded via the use of the removable media to provide intelligent services based on the analytics.

A method and apparatus of a device that uses a video jitter buffer to buffer the video frames for a received video stream is described. The device uses the video jitter buffer to estimate the delay variability of the frames and pick a target delay that will help harmonize the playback with minimal loss and delay. This is achieved by estimating the delay spread, which is the result of performing statistical analysis of the measured lags of the video frames received. The video jitter buffer provides target delay recommendation and reference frame information to the video player in order to anchor the playback and schedule the frames presentation time.

Disclosed herein are a method and apparatus for measuring video quality based on a perceptually sensitive region. The quality of video may be measured based on a perceptually sensitive region and a change in the perceptually sensitive region. The perceptually sensitive region includes a spatial perceptually sensitive region, a temporal perceptually sensitive region, and a spatio-temporal perceptually sensitive region. Perceptual weights are applied to a detected perceptually sensitive region and a change in the detected perceptually sensitive region. Distortion is calculated based on the perceptually sensitive region and the change in the perceptually sensitive region, and a result of quality measurement for a video is generated based on the calculated distortion.

A self-calibrating display can detect and compensate for binocular disparity or other visual imperfection of the display. The display includes a pair of projection units for projecting test light carrying test images through waveguides, which are normally used to carry images to left and right eyes of a user. A detection unit detects the test light propagated through the waveguides, and extracts the test images. Position of reference features in the detected test images may be used to determine binocular disparity, and luminance and color distribution across the test images may be used to determine the illumination and color uniformity of the images displayed to the user. After the visual defects have been detected, they may be reduced or compensated for by pre-emphasizing or shifting images to be displayed to the left and right eyes of the user.

In one embodiment, a method is provided. The method includes comparing a first hue of a displayed test pattern with a second hue of a generated test pattern to determine one or more differences in hue. The displayed test pattern is displayed on a display of a media viewing device, and the media viewing device receives and displays digital content. The displayed test pattern includes a first plurality of steps, and a first step of the first plurality of steps includes the first hue. The generated test pattern includes a second plurality of steps, and each step of the second plurality of steps includes the second hue. The one or more differences in hue are utilized to determine a quality of the digital content displayed on the media viewing device.

The present invention relates to a power and video redundancy system for a display system of a smartboard. More particularly, the present invention relates to a power and video redundancy system applied to a smartboard display system which minimizes the user's inconvenience due to the failure or damage of components and enables the manager to repair or change the parts without the user being aware of the loss or damage.

A video quality assessment method and device are provided. The video quality assessment method includes: obtaining a to-be-assessed video, where the to-be-assessed video includes a forward error correction (FEC) redundancy data packet; when a quantity of lost data packets of a first source block in the to-be-assessed video is less than or equal to a quantity of FEC redundancy data packets of the first source block, generating a first summary packet for a non-lost data packet of the first source block, and generating a second summary packet for a lost data packet of the first source block; and calculating a mean opinion score of video (MOSV) of the to-be-assessed video based on the first summary packet and the second summary packet. The MOSV calculated according to the method is more consistent with real video experience of a user, so accuracy of video quality assessment can be improved.

Aspects of the present disclosure relate to the assessment on how well and consistent the quality of a video asset satisfies a given target quality level or a time-varying target quality curve, and the optimization on encoding configuration to achieve the best compromise between satisfying the target quality requirement and saving the bit rate/bandwidth cost. The application scope of the present disclosure is generally in, but not limited to, the field of video coding and distributions, including both live and file-based video encoding, broadcasting and streaming systems. Methods and systems implemented based on the present disclosure may achieve the highest accuracy approaching any given target quality with the smoothest quality variation over time, while maximally reduce bit rate/bandwidth and video distribution cost by optimally determining video encoding configurations and parameters.

In an arrangement where a physical phenomenon affects a digital video camera and is measured or sensed by a sensor, a delay of a digital video stream from the digital video camera is estimated. The digital video stream is processed by a video processor for producing a signal that represents the changing over time of the effect of the physical phenomenon on the digital video camera. The signal is then compared with the sensor output signal, such as by using cross-correlation or cross-convolution, for estimating the time delay between the compared signals. The estimated time delay may be used for synchronizing when combining additional varied data to the digital video stream for low-error time alignment. The physical phenomenon may be based on mechanical position or motion, such as pitch, yaw, or roll. The time delay estimating may be performed once, upon user control, periodically, or continuously.

A method for video quality assessment implemented by a computing device, where the method includes selecting a first set of frames from an input video source, determining features of the first set of frames for input into a first machine learning model, applying the first machine learning model to the features of the first set of frames to obtain a video quality score and a confidence score, and triggering a retraining of the first machine learning model in response to the confidence score being determined to be out of bounds.