Microsoft technology licensing, llc (20240184852). NEURAL NETWORK TARGET FEATURE DETECTION simplified abstract

NEURAL NETWORK TARGET FEATURE DETECTION

Organization Name

microsoft technology licensing, llc

Inventor(s)

Hamidreza Vaezi Joze of Redmond WA (US)

Vivek Pradeep of Redmond WA (US)

Karthik Vijayan of Berkshire (GB)

NEURAL NETWORK TARGET FEATURE DETECTION - A simplified explanation of the abstract

This abstract first appeared for US patent application 20240184852 titled 'NEURAL NETWORK TARGET FEATURE DETECTION

Simplified Explanation

The patent application describes a method for training a neural network to detect target features in images. Labeled images are used to train the network, which is then tested on unlabeled images to identify target features.

• The neural network is trained using a first data set of labeled images.
• Labeled images are divided into tiles, and feature anchors are generated for each tile.
• Target features are detected in a second data set of unlabeled images.
• Images with undetected target features are labeled.
• A third data set is generated with the first data set and labeled images from the second data set.
• The neural network is trained using the third data set.

Key Features and Innovation

• Training a neural network to detect target features in images.
• Using labeled images to train the network and testing it on unlabeled images.
• Dividing images into tiles and generating feature anchors for each tile.
• Detecting target features in unlabeled images and labeling images with undetected features.
• Generating a new data set for training the neural network.

Potential Applications

This technology can be applied in various fields such as:

• Image recognition
• Object detection
• Medical imaging
• Surveillance systems

Problems Solved

• Efficiently detecting target features in images
• Improving the accuracy of image recognition systems
• Automating the process of labeling images with target features

Benefits

• Enhanced image analysis capabilities
• Increased accuracy in detecting specific features
• Streamlined training process for neural networks

Commercial Applications

• "Enhanced Image Recognition Technology for Target Feature Detection"
• This technology can be utilized in industries such as:
  • Healthcare for medical imaging analysis
  • Security for surveillance systems
  • Manufacturing for quality control processes

Prior Art

Information on prior art related to this technology is not provided.

Frequently Updated Research

There is no information on frequently updated research related to this technology.

Questions about Neural Network Training for Image Feature Detection

Question 1

How does the neural network differentiate between different types of target features in images?

The neural network uses feature anchors generated for each tile to identify specific target features within the image.

Question 2

What is the significance of dividing images into tiles for training the neural network?

Dividing images into tiles helps the neural network focus on specific regions of the image, improving the accuracy of feature detection.

Original Abstract Submitted

a method of training a neural network for detecting target features in images is described. the neural network is trained using a first data set that includes labeled images, where at least some of the labeled images having subjects with labeled features, including: dividing each of the labeled images of the first data set into a respective plurality of tiles, and generating, for each of the plurality of tiles, a plurality of feature anchors that indicate target features within the corresponding tile. target features that correspond to the plurality of feature anchors are detected in a second data set of unlabeled images. images of the second data set having target features that were not detected are labeled. a third data set that includes the first data set and the labeled images of the second data set is generated. the neural network is trained using the third data set.