Adaptive input sampling for machine learning

Adaptive input sampling for machine-learned models optimizes frame rates and resolutions based on video content analysis, addressing inefficiencies in existing systems by reducing costs and improving accuracy.

WO2026128235A1 Publication Date: 2026-06-18GDM HOLDING LLC
0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
GDM HOLDING LLC
Filing Date
2025-11-26
Publication Date
2026-06-18

AI Technical Summary

⚠Technical Problem

Existing machine-learning systems face inefficiencies in processing video data due to fixed frame rates and resolutions, leading to increased computational and communication costs without ensuring optimal input quality for inference tasks.

⚗Method used

Adaptive input sampling for machine-learned models that dynamically adjust frame rates and resolutions based on video content analysis, using machine-learned and non-machine-learned metrics to prioritize high-value frames for processing.

🎯Benefits of technology

Reduces computational and communication costs while improving inference accuracy by selectively sampling frames based on their relevance, enhancing performance under resource constraints.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
Patent Text Reader

Abstract

Systems and methods are provided. An example method can include providing, by a computing system comprising one or more computing devices, to a first machine-learned model, one or more first video frames. The example method can include determining, by the computing system based at least in part on the one or more first video frames, whether to provide one or more second video frames to the first machine-learned model. The example method can include providing, by the computing system responsive to determining that the one or more second video frames should be provided to the first machine-learned model, the one or more second video frames to the first machine-learned model. The example method can include generating, by the first machine-learned model based at least in part on the one or more second video frames, an output
Need to check novelty before this filing date? Find Prior Art