Watermarking synchronized inputs for machine learning
AssigneeSony Interactive Entertainment Inc.
InventorMatthew Bennett
Life of this patent
Estimating the term…
Abstract
A method and system for providing synchronized input feedback, comprising receiving an input event, encoding the input event in an output stream wherein the encoding of the input event is synchronized to a specific event and reproducing the output stream through an output device whereby the encoded input event in the reproduced output stream is imperceptible to the user.
Claims19
A method for providing synchronized input feedback:
- a) receiving an input event;
- b) encoding the input event in an output stream wherein the encoding of the input event is synchronized to a specific event occurring within a videogame, wherein the input event corresponds to user operation of a controller, wherein the output stream includes an audio stream and wherein a different length of the input event is encoded by a different length of a tone in the audio stream and a different type of the input event is encoded by a different frequency tone in the audio stream;
- c) coupling the output stream to one or more neural networks, wherein the encoded input event in the output stream is Imperceptible to the user when reproduced through an output device.
A non-transitory computer readable medium comprising computer-readable instructions embedded thereon that when executed by a computer cause the computer to enact the method comprising:
- a) receiving an input event;
- b) encoding the input event in an output stream wherein the encoding of the input event is synchronized to a specific event occurring within a videogame, wherein the input event corresponds to user operation of a controller, wherein the output stream includes an audio stream and wherein a different length of the input event is encoded by a different length of a tone in the audio stream and a different type of the input event is encoded by a different frequency tone in the audio stream;
- c) coupling the output stream to one or more neural networks wherein the encoded input event in the output stream is undetectable to the user when reproduced through an output device.
A processor;
- memory;
- non-transitory instructions in the memory that when executed cause the processor to enact the method comprising:
- a) receiving an input event;
- b) encoding the input event in an output stream wherein the encoding of the input event is synchronized to a specific event occurring within a videogame, wherein the input event corresponds to user operation of a controller, wherein the output stream Includes an audio stream and wherein a different length of the input event is encoded by a different length of a tone in the audio stream and a different type of the input event is encoded by a different frequency tone in the audio stream;
- c) coupling the output stream to one or more neural networks, wherein the encoded input event in the output stream is undetectable to the user when reproduced through an output device.
A method for improved machine learning training comprising;
- a) receiving an output stream comprising an encoded input event synchronized to a specific event occurring within a videogame, wherein the input event corresponds to user operation of a controller, wherein the encoded input event is undetectable to the user, wherein the encoded input event in the reproduced output stream is decodable for use by one or more neural networks, and wherein the output stream Includes an audio stream and wherein a different length of the input event is encoded by a different length of a tone in the audio stream and a different type of the input event is encoded by a different frequency tone in the audio stream;
- b) filtering the output stream to recover the encoded input event wherein the encoded input event is synchronized to a specific event;
- c) training a neural network to associate the encoded input event with the specific event.
Dependencies are inferred from the claim text.