Neural Network Training Using Partial Datasets and User Weights
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing AI systems require large datasets for effective neural network training, leading to inefficiencies in category recognition and performance improvement, especially when dealing with small datasets.
Innovation Solution
A neural network training method that utilizes a category-specified partial dataset to recognize categories in other datasets, refining the model through iterative training with recognized images, and incorporating user-preferred image weights or inputs to enhance recognition accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a neural network model is trained using a large dataset, then recognition accuracy is improved, but training time and computational resources increase
Solution Approach 1:
The patent applies preliminary action by pre-training the neural network model using a category-specified partial dataset before deploying it for recognition tasks. This pre-training phase prepares the model in advance with essential category knowledge, enabling it to achieve good recognition accuracy on small datasets without requiring extensive training time on large datasets during deployment.
Solution Approach 2:
The patent uses partial action by training the neural network model only on a category-specified partial dataset rather than requiring a complete large dataset. The model is trained on images belonging to specific categories of interest, which is sufficient for achieving accurate recognition in those categories while significantly reducing training time and computational resources compared to training on all possible categories.
2Productivity
If a neural network model is trained on a small dataset, then training efficiency is improved, but recognition accuracy deteriorates
Solution Approach 1:
The patent applies preliminary action by pre-training the neural network model using a category-specified partial dataset before deploying it for recognition tasks. This pre-training phase prepares the model in advance with essential category knowledge, enabling it to achieve good recognition accuracy on small datasets without requiring extensive training time on large datasets during deployment.
3Measurement precision
If user-preferred image weights are incorporated into training, then recognition accuracy for preferred categories is improved, but model complexity increases
Solution Approach 1:
The patent applies local quality by assigning different weights to different categories based on user preferences. Instead of uniformly treating all categories, the model applies higher weights to user-preferred categories during training, allowing it to achieve higher recognition accuracy for those specific categories while maintaining simpler processing for other categories. This selective weighting approach enhances performance where needed without uniformly increasing model complexity.
Data Source
AI summary
Provided are an AI system for simulating functions such as recognition, determination, and so forth of human brains by using a mechanical learning algorithm like deep learning, or the like, and an application thereof. In particular, according to the AI system and the application thereof, a neural network training method includes obtaining a plurality of first images belonging to a particular category and a plurality of second images for which a category is not specified, training a neural network model for category recognition, based on the plurality of first images belonging to the particular category, recognizing at least one second image corresponding to the particular category among the plurality of second images, by using the trained neural network model, and modifying and refining the trained neural network model based on the recognized at least one second image.