Landscape Image Estimation via Deep Learning Models

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Users with ambiguous requests for products that blend with a landscape image face challenges in finding suitable products, as existing systems struggle to estimate the image represented by a captured landscape effectively.

Innovation Solution

An information processing apparatus and program that uses machine learning models, specifically deep learning techniques, to estimate the image, place, and color of a landscape from input pictures, allowing for the identification of matching products based on these attributes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If machine learning models are used to estimate landscape image attributes, then measurement precision of landscape attributes is improved, but device complexity increases

Engineering Contradiction:

Improvemeasurement precisionVSAvoiddevice complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

A machine learning model serves as an intermediary component that bridges the input picture and the estimated landscape attributes. The model processes the input image and outputs estimated attributes such as image type, place, and color, enabling accurate measurement without requiring complex manual analysis systems.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces traditional mechanical or manual methods of landscape analysis with a machine learning-based system. Instead of using complex rule-based engines or manual interpretation, the system uses trained neural networks to automatically estimate landscape attributes from input pictures, simplifying the overall system architecture while improving precision.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If deep learning techniques are applied to estimate landscape attributes, then reliability of product recommendation is improved, but loss of energy increases

Engineering Contradiction:

ImprovereliabilityVSAvoiduse of energy

Core Design Contradiction:

ReliabilityVSUse of energy by moving object

Solution Approach 1:

The machine learning model is trained in advance on large datasets of landscape images and their corresponding attributes. This preliminary training phase allows the model to learn complex patterns and relationships, enabling it to provide reliable product recommendations with minimal computational effort during actual use, thereby reducing energy consumption during operation.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system adjusts various parameters of the machine learning model, such as model architecture, learning rate, and training data composition, to optimize the balance between recommendation reliability and energy consumption. By fine-tuning these parameters, the system achieves high accuracy while minimizing the computational resources required during inference.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS11232320B2Information processing apparatus and non-transitory computer readable medium storing information processing program for estimating image represented by captured landscape

Publication Date: 2022.01.25 FUJIFILM BUSINESS INNOVATION CORP

AI summary

An information processing apparatus includes a receiving unit that receives an input picture obtained by capturing a landscape including an object, and an estimation unit that estimates an image represented by the landscape appearing in the input picture based on a learning model in which the input picture received by the receiving unit is input.