Back to results
US 2026/0154955 A1Published application
Instruction generation using one or more neural networks
AssigneeNVIDIA Corporation
InventorsPranit P. Kothari, Siddhant Pardeshi, Vinayak Vilas Gaikwad
Life of this patent
Estimating the term…
Abstract
Apparatuses, systems, and techniques are presented for generating instructional text. In at least one embodiment, an instructional video is analyzed to determine logical steps of a process or task demonstrated in that video, and instructive text is generated for those logical steps.
Claims21
1-30. (canceled)
One or more processors, comprising:
- circuitry to use one or more neural networks to generate one or more procedures by:
- receiving one or more video frames depicting one or more series of actions;
- generating, using the one or more neural networks, one or more contextualized representations of one or more actions of the one or more series of actions based, at least in part, on a portion of the one or more video frames, the portion depicting the one or more actions;
- and
- generating one or more portions of text based, at least in part, on the contextualized representations and one or more video frames.
2 more elements
A system, comprising:
- one or more processors and memory storing executable instructions that, when performed by the one or more processors in response to a call to an application programming interface (API), use one or more neural networks to generate one or more procedures by:
- receiving one or more video frames depicting one or more series of actions;
- generating, using the one or more neural networks, one or more contextualized representations of one or more actions of the one or more series of actions based, at least in part, on a portion of the one or more video frames, the portion depicting the one or more actions;
- and
- generating one or more portions of text based, at least in part, on the contextualized representations and the one or more video frames.
2 more elements
A method, comprising:
- using one or more neural networks to generate one or more procedures by:
- receiving one or more video frames depicting one or more series of actions;
- generating, using the one or more neural networks, one or more contextualized representations of one or more actions of the one or more series of actions based, at least in part, on a portion of the one or more video frames, the portion depicting the one or more actions;
- and
- generating one or more portions of text based, at least in part, on the contextualized representations and the one or more video frames.
2 more elements
Dependencies are inferred from the claim text.