# US 2026/0205558 A1 · Video prediction using one or more neural networks

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2026/0205558 A1
- Document ID: US20260205558A1
- Status: Published application
- Kind: A1
- Application number: 19/559,798
- Filed: 2026-03-06
- Published: 2026-07-16
- Earliest claimed: 2019-09-03
- Assignees: NVIDIA Corporation
- Applicants: NVIDIA Corporation
- Inventors: Kevin Shih; Aysegul Dundar; Animesh Garg; Robert Pottorff; Andrew Tao; Bryan Catanzaro
- Web page: https://patentgrep.com/patent/US20260205558A1

## CPC

- H04N 7/0135: {involving interpolation processes (interpolation-based image scaling G06T3/4007; interpolation for video coding H04N19/587, H04N19/59)}
- G06F 18/214: Generating training patterns; Bootstrap methods, e.g. bagging or boosting
- G06F 18/217: {Validation; Performance evaluation; Active pattern learning techniques}
- G06N 3/044: Recurrent networks, e.g. Hopfield networks
- G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
- G06N 3/045: Combinations of networks
- G06N 3/0455: Auto-encoder networks; Encoder-decoder networks
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/08: Learning methods
- G06N 3/084: Backpropagation, e.g. using gradient descent
- G06N 3/0895: Weakly supervised learning, e.g. semi-supervised or self-supervised learning
- G06N 3/09: Supervised learning
- G06N 3/094: Adversarial learning
- G06T 9/002: {using neural networks}
- G06N 3/047: Probabilistic or stochastic networks
- G06N 3/049: Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
- G06N 3/063: using electronic means
- G06N 3/088: Non-supervised learning, e.g. competitive learning
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL

## Abstract

Apparatuses, systems, and techniques to enhance video are disclosed. In at least one embodiment, one or more neural networks are used to create, from a first video, a second video having one or more additional video frames.

## Claims

### Claim 1

A processor comprising:
one or more arithmetic logic units (ALUs) to be configured to infer frames for a video using one or more neural networks trained using one or more temporal pose representations.

### Claim 2 (depends on 1)

The processor of claim 1, wherein video frames used to train the one or more neural networks are used to generate the one or more temporal pose representations and a time-invariant appearance representation.

### Claim 3 (depends on 2)

The processor of claim 2, wherein color jittering and thin-plate-spline (TPS) warping enforce appearance invariance and localization properties for the one or more temporal pose representations.

### Claim 4 (depends on 2)

The processor of claim 2, wherein the one or more temporal pose representations comprise features represented by Gaussian heat maps and parameterized as a mean and a covariance.

### Claim 5 (depends on 2)

The processor of claim 2, wherein the one or more temporal pose representations and the time-invariant appearance representation are used to reconstruct an input video frame, and wherein a loss value resulting from comparing the input video frame with the reconstructed video frame is minimized by adjusting one or more network parameters of the one or more neural networks.

### Claim 6 (depends on 1)

The processor of claim 1, wherein motion of the one or more temporal pose representations is modeled over time using a temporal encoder including one or more long short-term memory (LSTM) networks.

### Claim 7 (depends on 1)

The processor of claim 1, wherein the inferred frames are used to generate a video with a higher frame rate, fewer dropped frames, or additional content.

### Claim 8

A system comprising:
one or more processors to be configured to infer frames for a video using one or more neural networks trained using one or more temporal pose representations; and
one or more memories to store the one or more neural networks.

### Claim 9 (depends on 8)

The system of claim 8, video frames used to train the one or more neural networks are used to generate a pose representation and a time-invariant appearance representation.

### Claim 10 (depends on 9)

The system of claim 9, wherein color jittering and thin-plate-spline (TPS) warping enforce appearance invariance and localization properties for the temporal pose representations.

### Claim 11 (depends on 9)

The system of claim 9, wherein the one or more temporal pose representations comprise features represented by Gaussian heat maps parameterized as a mean and covariance.

### Claim 12 (depends on 9)

The system of claim 9, wherein the pose representation and the time-invariant appearance representation are used to reconstruct an input video frame, and wherein a loss value resulting from comparing the input video frame with the reconstructed video frame is minimized by adjusting one or more network parameters of the one or more neural networks.

### Claim 13 (depends on 8)

The system of claim 8, wherein motion of the one or more temporal pose representations is modeled over time using a temporal encoder including one or more long short-term memory (LSTM) networks.

### Claim 14

A machine-readable medium having stored thereon a set of instructions, which if performed by one or more processors, cause the one or more processors to at least:
infer frames for a video using one or more neural networks trained using one or more temporal pose representations.

### Claim 15 (depends on 14)

The machine-readable medium of claim 14, wherein video frames used to train the one or more neural networks are used to generate a pose representation and a time-invariant appearance representation.

### Claim 16 (depends on 15)

The machine-readable medium of claim 15, wherein color jittering and thin-plate-spline (TPS) warping enforce appearance invariance and localization properties for the temporal pose representations.

### Claim 17 (depends on 15)

The machine-readable medium of claim 15, the one or more temporal pose representations comprise features represented by Gaussian heat maps parameterized as a mean and covariance.

### Claim 18 (depends on 15)

The machine-readable medium of claim 15, wherein the pose representation and the time-invariant appearance representation are used to reconstruct an input video frame, and wherein a loss value resulting from comparing the input video frame with the reconstructed video frame is minimized by adjusting one or more network parameters of the one or more neural networks.

### Claim 19 (depends on 14)

The machine-readable medium of claim 14, wherein motion of the one or more temporal pose representations is modeled over time using a temporal encoder including one or more long short-term memory (LSTM) networks.
