Electronic device, method, and non-transitory computer-readable storage medium for determining part of image and / or video to be changed on basis of natural language input

The electronic device addresses the challenge of editing images and videos using natural language inputs by employing AI models for object detection and user interfaces, enabling precise subject region modification and improved user interaction.

WO2026142403A1PCT designated stage Publication Date: 2026-07-02SAMSUNG ELECTRONICS CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
SAMSUNG ELECTRONICS CO LTD
Filing Date
2025-11-12
Publication Date
2026-07-02

AI Technical Summary

Technical Problem

Existing electronic devices struggle to efficiently edit images and videos based on natural language input, particularly in determining and modifying specific subject regions, due to limitations in user interface and artificial intelligence capabilities.

Method used

An electronic device equipped with a processor and memory, utilizing AI models for object detection, speech-to-text conversion, and image editing, allows users to input natural language commands to select and modify subject regions within images and videos, with the aid of user interfaces for precise selection.

Benefits of technology

Enables seamless editing of images and videos by interpreting natural language inputs, allowing users to accurately identify and modify subject regions, enhancing user interaction and functionality in devices like TVs and set-top boxes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure KR2025095704_02072026_PF_FP_ABST
    Figure KR2025095704_02072026_PF_FP_ABST
Patent Text Reader

Abstract

The present disclosure relates to an artificial intelligence (AI) system utilizing a machine learning algorithm, and an application thereof. An electronic device according to one embodiment may acquire, on the basis of a user input related to editing of an image, regions related to a plurality of subjects, corresponding to a type related to the user input. The electronic device may control an output of at least one user interface (UI) for selecting at least one of the regions included in the image and at least partially overlapping each other, on the basis of acquiring the regions related to the plurality of subjects.
Need to check novelty before this filing date? Find Prior Art