# BWD-DETR: A Robust Framework for Bright-Field Wafer Defect Detection

## TL;DR

This paper introduces BWD-DETR, a new framework that improves the detection of tiny defects on patterned wafers using bright-field imaging.

## Contribution

BWD-DETR integrates a wavelet backbone and novel modules to enhance sub-micron defect detection accuracy in bright-field wafer imaging.

## Key findings

- BWD-DETR achieves an AP50 of 96.56% and AP50:95 of 54.94% in bright-field wafer defect detection.
- The framework outperforms the baseline by 1.64% and 2.17% in AP50 and AP50:95 metrics respectively.
- The proposed method effectively enhances detection of sub-micron defects on wafer surfaces.

## Abstract

Optical defect detection based on bright-field imaging is currently one of the most widely applied inspection techniques in wafer fabrication. However, particle defects on the surface of patterned wafers are often small in size. Under bright-field optical imaging conditions, defect signals are easily overwhelmed by complex background textures and noise, seriously affecting the detectability and positioning accuracy of defects. To address this issue, this paper proposes BWD-DETR, a detection framework tailored for wafer surface defects under bright-field imaging. Based on the RT-DETR baseline, this framework integrates a wavelet backbone, an SMFI module, and a CAS-Fusion module, achieving an AP50 of 96.56% and an AP50:95 of 54.94% in bright-field wafer defect detection, with improvements of 1.64% and 2.17% over the baseline, respectively. The proposed method can effectively enhance the detection capability for sub-micron defects on the wafer surface.