Virtual Viewpoint Path Generation Using Learned Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users face significant time and effort in designating virtual viewpoints for generating virtual viewpoint images, and pre-defined viewpoints may not be suitable for the scene, leading to inefficient image processing.
Innovation Solution
An information processing apparatus that uses a learned model to generate virtual viewpoint path data by obtaining object position information from multi-viewpoint image data, reducing the need for user input and improving viewpoint selection through automatic path generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the virtual viewpoint is designated based on user input every time, then the virtual viewpoint can be customized for each generation, but it takes the user great time and effort
Solution Approach 1:
The system pre-generates multiple candidate virtual viewpoint paths before the user needs to select one. These candidate paths are stored and can be quickly retrieved and selected based on user feedback or automatic selection, eliminating the need for time-consuming manual designation each time a virtual viewpoint image is generated.
Solution Approach 2:
The system creates multiple copies of potential virtual viewpoint paths in advance. Instead of generating one viewpoint at a time through manual input, multiple candidate paths are prepared beforehand, allowing the user to simply select from pre-generated options rather than creating each viewpoint from scratch.
2Loss of time
If the virtual viewpoint is fixedly defined in advance, then user effort is reduced, but there is a possibility that a virtual viewpoint not suitable for a scene is set
Solution Approach 1:
The system provides multiple candidate virtual viewpoint paths that are dynamically selectable based on the specific scene being captured. Rather than using a single fixed viewpoint, the system adapts by offering multiple pre-generated options tailored to different scene characteristics, allowing selection that best fits the current situation.
Solution Approach 2:
The system varies parameters of virtual viewpoint paths (such as position, angle, and trajectory) to create multiple candidate paths suitable for different scene types. By changing these parameters in advance to cover various scenarios, the system ensures that an appropriate viewpoint is available for any given scene without requiring manual adjustment.
3Measurement precision
If manual virtual viewpoint designation is used, then viewpoint accuracy can be controlled, but productivity is reduced due to repeated user input
Solution Approach 1:
Multiple candidate virtual viewpoint paths are generated in advance with high precision through automated algorithms. These pre-computed paths maintain accurate viewpoint designation while eliminating the need for repeated manual input, thereby improving productivity without sacrificing accuracy.
Solution Approach 2:
The system performs automatic generation of candidate virtual viewpoint paths using algorithms that analyze scene characteristics and generate appropriate viewpoints autonomously. This self-service capability maintains viewpoint accuracy while freeing users from manual designation tasks, thus improving overall efficiency.
Data Source
AI summary
The technology disclosed herein is an information processing apparatus comprising: one or more memories storing instructions; and one or more processors executing the instructions to function as: an obtaining unit configured to obtain information for specifying a position of an object included in multi-viewpoint image data obtained by image capturing using a plurality of imaging apparatuses; and a generation unit configured to generate a virtual viewpoint path data to generate virtual viewpoint image data by inputting the information obtained by the obtaining unit to an output unit which is a learned model learned from the virtual viewpoint path data to be training data and at least information for specifying a position of an object to be input data corresponding to the virtual viewpoint path data and is configured to output virtual viewpoint data by receiving input of information for specifying a position of an object.