Machine-Learned Models for Imperceptible Message Watermarking in Videos

A computer-implemented method for training a machine-learned watermark encoding model, the method comprising:

• obtaining a message vector and video data comprising a plurality of video frames;
• processing the message vector and the video data with the machine-learned watermark encoding model to obtain encoded video data comprising a plurality of encoded video frames, wherein at least one of the plurality of encoded video frames comprises spatial-temporal watermark encoding data descriptive of the message vector;
• processing the encoded video data with a differentiable distortion layer to apply one or more distortions to the encoded video data in a differentiable fashion;
• and
• processing the encoded video data, subsequent to the application of the one or more distortions, with a machine-learned watermark decoding model to obtain a reconstructed message vector.

A computer-implemented method for training a machine-learned watermark encoding model, the method comprising:

• obtaining a message vector and video data comprising a plurality of video frames;
• processing the message vector and the video data with the machine-learned watermark encoding model to obtain encoded video data including spatial-temporal watermark encoding data descriptive of the message vector;
• applying one or more distortions to the encoded video data via a differentiable distortion layer;
• and
• decoding a reconstructed message vector from encoded video data that is distorted by processing the encoded video data with a decoder portion of a machine-learned watermark decoding model to obtain and weight a plurality of feature encodings having different spatial-temporal dimensions.

A computing system comprising:

• one or more processors;
• and
• one or more non-transitory computer-readable media storing instructions that, when executed by the one or more processors, cause the computing system to perform operations, the operations comprising:
• obtaining a message vector and video data comprising a plurality of video frames;
• processing the message vector and the video data with a machine-learned watermark encoding model to obtain encoded video data comprising a plurality of encoded video frames, wherein at least one of the plurality of encoded video frames comprises spatial-temporal watermark encoding data descriptive of the message vector;
• processing the encoded video data with a differentiable distortion layer to apply one or more distortions to the encoded video data in a differentiable fashion;
• and
• processing the encoded video data, subsequent to the application of the one or more distortions, with a machine-learned watermark decoding model to obtain a reconstructed message vector.