Latent Space Artifact Vector Subtraction for Semiconductor Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In semiconductor substrate imaging, artifacts such as noise and charging introduce errors in feature measurement, which are critical for accurate lithographic processes due to stringent tolerances in edge placement and critical dimension variations.
Innovation Solution
A method involving encoding images into a latent space to identify and subtract artifact vectors, thereby decoding images with reduced or removed artifacts, ensuring more accurate feature measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional imaging methods are used to capture semiconductor substrate features, then the imaging process is simple and fast, but artifacts such as noise and charging introduce measurement errors that reduce measurement precision
Solution Approach 1:
The patent segments the image processing task into distinct components: encoding the input image into a latent space representation, identifying and subtracting artifact vectors from the encoding, and decoding the cleaned encoding to produce the final image. This segmentation allows artifact removal to be performed as a separate, targeted operation that improves measurement precision without overwhelming complexity
Solution Approach 2:
The patent introduces a latent space encoding as an intermediary representation between the raw image and the final processed image. This intermediary encoding serves as a mediator where artifact vectors can be identified and subtracted in a transformed domain, making the artifact removal process more effective and manageable than direct manipulation in the original image space
2Measurement precision
If artifact removal processing is applied to enhance measurement accuracy, then measurement precision improves, but processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary encoding of the image into latent space representation before artifact removal operations. This preliminary transformation prepares the data in an optimized format that facilitates more efficient artifact identification and subtraction, reducing the computational burden of subsequent processing steps compared to working with raw pixel data
Solution Approach 2:
The patent changes the representation parameters of the image data by transforming it into a latent space encoding. This parameter change from pixel space to latent space enables more efficient manipulation and processing of the image data, allowing artifact removal to be performed with reduced computational complexity and faster processing time
3Reliability
If conventional image analysis is used without artifact removal, then processing is faster and simpler, but measurement accuracy and reliability are compromised due to artifacts
Solution Approach 1:
The patent extracts artifact components from the image encoding by identifying and subtracting artifact vectors from the latent space representation. This extraction isolates the harmful artifact elements from the useful image information, allowing reliable measurements to be obtained from the cleaned encoding without requiring complete redesign of the entire imaging system
Data Source
AI summary
An inspection tool comprises an imaging system configured to image a portion of a semiconductor substrate. The inspection tool may further comprise an image analysis system configured to obtain an image of a structure on the semiconductor substrate from the imaging system, encode the image of the structure into a latent space thereby forming a first encoding. the image analysis system may subtract an artifact vector, representative of an artifact in the image, from the encoding thereby forming a second encoding; and decode the second encoding to obtain a decoded image.