Nvidia Corporation

Pratyush Mahapatra

Ruthie D Lyle

Somayyeh Rahimi

Systems and methods of selectively compressing video data are disclosed. The proposed systems provide a computer-implemented process configured to classify a person's behavior(s) during a video and encode the behaviors as a representation of the video. The encoding will be tailor-generated based on the specific display configuration of the target device at which playback is expected to occur. Target device displays with lower resolution and video quality characteristics will trigger an encoding of the video data that has less complexity than target device displays with higher resolution and video quality characteristics. When playback of the video is requested at the target device, a reconstruction of the video is generated by a video synthesizer based on a reference image of the person and the encoding rather than the original video file, significantly reducing memory, processing, power, and bandwidth requirements.

Selective data encoding and machine learning video synthesis for content streaming systems and applications

Systems and methods of compressing video content as encoded data and selectively reconstructing portions of the content are disclosed. The proposed systems provide a computer-implemented process configured to classify a person's behavior(s) during a video and encode the behaviors as a representation of the video. When playback of the video is requested, a video navigation assistant will allow the end-user to select specific segments of the video based on topics discussed in the video and the codes that were generated to represent the video. The user is then able to move through segments of the video in a sequence that aligns with their viewing preferences.

Conversational AI-encoded language for video navigation

Systems and methods of compressing video data are disclosed. The proposed systems provide a computer-implemented process configured to classify a person's behavior(s) during a video and encode the behaviors as a representation of the video. When playback of the video is requested, a reconstruction of the video is generated by a video synthesizer based on a reference image of the person and the sequence of codes corresponding to their behavior during the video. Storage and transmission of the video can then be limited to the reference image and the behavioral codes rather than the video file itself, significantly reducing memory and bandwidth requirements.

Growing community of inventors

Santa Clara, CA, United States of America

Pratyush Mahapatra