Systems, methods, and storage media for selecting video portions for a video synopsis of streaming video content

A computer implemented method for selecting video portions synopsis of high viewer interest from streaming video content, the method comprising:

• extracting at least a portion of an audio track from a live stream of video content over time to create an audio file;
• converting the audio file from a time domain to a frequency domain to create a frequency domain audio signal as a superposition of sinusoids of varying magnitudes, frequencies, and phase offsets;
• generating a representation of the spectrum of frequencies of the frequency domain audio signal as it varies with time;
• applying a domain specific neural network image classification algorithm to the representation to generate interest probability scores for portions of the audio signal;
• selecting portions of the audio signal that meet or exceed a threshold probability score;
• correlating the selected portions of the audio signal to corresponding segments of the video content that has been streamed;
• and
• marking the corresponding segments of the video content as being of high viewer interest.

A non-transient computer-readable storage medium having instructions embodied thereon, the instructions being executable by one or more processors to perform a method for selecting video portions of high viewer interest from streaming video content, the method comprising:

• extracting at least a portion of an audio track from a live stream of video content over time to create an audio file;
• converting the audio file from a time domain to a frequency domain to create a frequency domain audio signal as a superposition of sinusoids of varying magnitudes, frequencies, and phase offsets;
• generating a representation of the spectrum of frequencies of the frequency domain audio signal as it varies with time;
• applying a domain specific neural network image classification algorithm to the representation to generate interest probability scores for portions of the audio signal;
• selecting portions of the audio signal that meet or exceed a threshold probability score;
• correlating the selected portions of the audio signal to corresponding segments of the video content that has been streamed;
• and
• marking the corresponding segments of the video content as being of high viewer interest.