Embodiments described herein include systems and methods for automatically creating compilation videos from an original video based on metadata associated with the original video. For example, a method for creating a compilation video may include determining a relevance score for video frames in an original video; selecting a plurality of relevant video frames from the original video based on the relevance score; selecting a plurality of video clips from the original video based on the relevance scores of the video frames; and creating a compilation video from the plurality of video clips. Each of the plurality of video clips, for example, may include at least one relevant video frame from the plurality of relevant video frames.

Techniques are disclosed for identifying and presenting video content that demonstrates features of a target product. The video content can be accessed, for example, from a media database of user-generated videos that demonstrate one or more features of the target product so that a user can see and hear the product in operation via a product webpage before making a purchasing decision. The product functioning videos supplement any static images of the target product and the textual product description to provide the user with additional context for each of the product's features, depending on the textual product description. The user can quickly and easily interact with the product webpage to access and playback the product functioning video to see and/or hear the product in operation.

A video content analysis system for mobile assets that includes track detection and infrastructure monitoring component. The track detection and infrastructure monitoring includes a reinforcement learning component, an object detection and location component, and an obstruction detection component to analyze video data, audio data, vehicle data, weather data, and route/manifest data to determine internal and/or external conditions relating to an the asset. A data acquisition and recording system uploads the data, internal and/or external condition information, object detection information, object location information, and obstruction detection information to a remote memory module and provides streaming video data in real-time to a remotely located user. Remotely located users can view the data in various view modes through a web browser or virtual reality device, which provides for quicker emergency response, validates the effectiveness of repairs and rerouting, and monitors crew performance and safety.

A method of selecting a frame from a plurality of video frames captured by a camera (120). The method determines features to which map points in a three dimensional space are projected. A histogram of the determined features for a plurality of regions in the frame is created. One of the regions may be determined as being an unmapped region based on the created histogram. The frame is selected based on the size of the unmapped region.

The disclosure includes a system and method for decomposing a video to salient fragments and synthesizing a video composition based on the salient fragments. A computer-implemented method receives a first set of salient fragments and a first set of clusters extracted from a video, where each cluster includes related salient fragments connected by a connectivity graph. The method determines a weight associated with each of the salient fragments and each of the clusters based on an activity level associated with the respective salient fragment or cluster and determine a permissible zone of activity. The method determines a spatial-temporal distortion to be applied to each salient fragment and cluster and synthesizes a video composition based on the first set of salient fragments, the first set of clusters and non-salient portions of the video using weighted editing.

Methods, systems, and apparatuses, among other things, may detect and store activity in videos based on a spatiotemporal graph representation. Spatiotemporal proximity graphs may be built based on one or more received tracks and may include one or more nodes and each node may include one or more attributes associated with a corresponding entity. One or more spatiotemporal relationships may be identified between the entities based on each spatiotemporal proximity graph one or more activities of the entities may be identified based on the spatiotemporal relationships.

An Internet of Things (IoT) or vehicle dash cam may store both a high-resolution and low-resolution video stream on a device. The video streams are selectively accessible by remote devices. Because of the relatively smaller storage requirements of low-resolution video files, retaining of additional video data on the vehicle device (beyond what would be possible with only high-resolution video) is possible. The user may be provided an option to adjust the amount of low-resolution and high-resolution video to store on the device. A combined media file may be generated by a device to include time-synced high-resolution video, low-resolution video, and/or metadata for a particular time period.

A method and apparatus for characteristic-based video processing include: in response to receiving a region of a picture of a video sequence, determining a characteristic in the region, the region being independent of other regions of the picture for video coding; determining a class associated with the region based on the characteristic, the class being selected from a plurality of classes; and encoding the region using a parameter set associated with the class, the parameter set being selected from a plurality of parameter sets for video coding at different quality levels.

Provided are an event storage device, an event search device, and an event alarm device for an event occurring in a surveillance area. The event storage device includes a processor and a memory and its operation is controlled by the processor. The processor is configured to receive images captured in a surveillance area, detect an event generated in the surveillance area, storing, in the memory, an image related to the event, among the captured images, and storing an event level of the detected event along with the stored image.

A retrieval device includes one or more processors. The processors acquire trajectory information indicating a movement trajectory of a target in time-series images. The processors acquire situation information indicating a peripheral situation of the target in the time-series images. The processors acquire a retrieval query containing a movement trajectory and a peripheral situation. The processors retrieve an image matching with the retrieval query among images contained in the time-series images based on the trajectory information and the situation information.