Methods, computer program products, and/or systems are provided that perform the following operations: obtaining video data; dividing the video data into a plurality of video fragments based, at least in part, on page detection; extracting one or more elements from each of the plurality of video fragments; determining element type data for each of the one or more extracted elements; generating element style data for the one or more extracted elements; determining page flow for the plurality of video fragments; and generating one or more pages based, at least in part, on the one or more elements extracted from the plurality of video fragments, the element type data, the element style data, and the page flow.

Apparatus, methods and systems of object recognition are disclosed. Embodiments of the inventive subject matter generates map-altered image data according to an object-specific metric map, derives a metric-based descriptor set by executing an image analysis algorithm on the map-altered image data, and retrieves digital content associated with a target object as a function of the metric-based descriptor set.

Aspects of the present disclosure involve a system comprising a medium storing a program and method for machine-learning based selection of a representative video frame. The program and method provide for receiving a set of video frames; determining a first subset of frames by removing frames outside of an image quality threshold; determining a second subset by removing frames outside of an image stillness threshold; computing feature data for each frame in the second subset; providing, for each frame in the second subset, the feature data to a machine learning model (MLM), the MLM being configured to output a score for each frame in the second subset of frames based on the feature data, the MLM having been trained with a first set of images labeled based on aesthetics, and with a second set of images labeled based on image quality; and selecting a frame based on output scores.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for partitioning videos. In one aspect, a method includes obtaining a partition of a video into one or more shots. Features are generated for each shot, including visual features and audio features. The generated features for each shot are provided as input to a partitioning neural network that is configured to process the generated features to generate a partitioning neural network output. The partition of the video into one or more chapters is determined based on the partitioning neural network output, where a chapter is a sequence of consecutive shots that are determined to be taken at one or more locations that are semantically related.

A method for pushing information is disclosed by which a client acquires the statistical characteristic information of a current video frame in real time during video playback on the client; the client then searches a first mapping relationship table consisting of mapping relations between the statistical characteristic information and index values that is established by the client for the index value that matches the acquired statistical characteristic information, and sends the index value thus found to a cloud server; the cloud server searches a second mapping relationship table consisting of mapping relations between the index values and push information that is established by the cloud server for the push information that corresponds to the index value; and finally the client receives and plays or displays the push information. There is also provided a system for pushing information.

A method of generating a lighting effect based on metadata of an audio stream, the method comprising steps of: extracting metadata items from the audio stream; retrieving a first set of one or more images based on the metadata items; controlling a light source to generate a lighting effect based on said first set of one or more images.

Apparatus, methods and systems of object recognition are disclosed. Embodiments of the inventive subject matter generates map-altered image data according to an object-specific metric map, derives a metric-based descriptor set by executing an image analysis algorithm on the map-altered image data, and retrieves digital content associated with a target object as a function of the metric-based descriptor set.

A processor receives a first list comprising a plurality of events from a portion of digital data of an unknown work and one or more metrics between each pair of adjacent events from the plurality of events. The processor compares the first list to a second list comprising events and metrics between events for a known work to determine a first quantity of hits and a second quantity of misses. The processor determines whether the first list matches the second list based on the first quantity of hits and the second quantity of misses. The processor determines that the unknown work is a copy of the known work responsive to determining that the first list matches the second list.

To generate a hash in video, a sample series of temporal difference samples is forming image order. A temporal averaging is performed and a rate of change detected to identify as distinctive events regions of high rate of change. Images having a distinctive event are labelled as distinctive images. For each image, the temporal spacing in images is calculated between that image other distinctive images to provide a set of temporal spacings for that image; and a hash is derived for that image from that set of temporal spacings.

Aspects of the present disclosure involve a system comprising a medium storing a program and method for machine-learning based selection of a representative video frame. The program and method provide for receiving a set of video frames; determining a first subset of frames by removing frames outside of an image quality threshold; determining a second subset by removing frames outside of an image stillness threshold; computing feature data for each frame in the second subset; providing, for each frame in the second subset, the feature data to a machine learning model (MLM), the MLM being configured to output a score for each frame in the second subset of frames based on the feature data, the MLM having been trained with a first set of images labeled based on aesthetics, and with a second set of images labeled based on image quality; and selecting a frame based on output scores.