The evolution of herbicide-resistant weed species is a serious threat for weed control. Therefore, we need an improved understanding of how gene regulation confers herbicide resistance in order to slow the evolution of resistance. The present study analyzed differentially expressed genes after glyphosate treatment on a glyphosate-resistant Tennessee ecotype (TNR) of horseweed (Conyza canadensis), compared to a susceptible biotype (TNS). A read size of 100.2 M was sequenced on the Illumina platform and subjected to de novo assembly, resulting in 77,072 gene-level contigs, of which 32,493 were uniquely annotated by a BlastX alignment of protein sequence similarity. The most differentially expressed genes were enriched in the gene ontology (GO) term of the transmembrane transport protein. In addition, fifteen upregulated genes were identified in TNR after glyphosate treatment but were not detected in TNS. Ten of these upregulated genes were transmembrane transporter or kinase receptor proteins. Therefore, a combination of changes in gene expression among transmembrane receptor and kinase receptor proteins may be important for endowing non-target-site glyphosate-resistant C. canadensis.
Anomaly detection and localization in industrial images are essential for automated quality inspection. PaDiM, a prominent method, models the distribution of normal image features extracted by pre-trained Convolutional Neural Networks (CNNs) but reduces dimensionality through random channel selection, potentially discarding structured information. We propose Wavelet-Enhanced PaDiM (WE-PaDiM), which integrates Discrete Wavelet Transform (DWT) analysis with multi-layer CNN features in a structured manner. WE-PaDiM applies 2D DWT to feature maps from multiple backbone layers, selects specific frequency subbands (e.g., LL, LH, HL), spatially aligns them, and concatenates them channel-wise before modeling with PaDiM's multivariate Gaussian framework. This DWT-before-concatenation strategy provides a principled method for feature selection based on frequency content relevant to anomalies, leveraging multi-scale wavelet information as an alternative to random selection. We evaluate WE-PaDiM on the challenging MVTec AD dataset with multiple backbones (ResNet-18 and EfficientNet B0-B6). The method achieves strong performance in anomaly detection and localization, yielding average results of 99.32% Image-AUC and 92.10% Pixel-AUC across 15 categories with per-class optimized configurations. Our analysis shows that wavelet choices affect performance trade-offs: simpler wavelets (e.g., Haar) with detail subbands (HL or LH/HL/HH) often enhance localization, while approximation bands (LL) improve image-level detection. WE-PaDiM thus offers a competitive and interpretable alternative to random feature selection in PaDiM, achieving robust results suitable for industrial inspection with comparable efficiency.