2026-10-14T13:11:49Z https://api.figshare.com/v2/oai

oai:figshare.com:article/33763979 2026-09-14T22:06:10Z category_146 category_15 category_19 category_21 category_873 category_734 category_122 portal_63 item_type_3 month_year_09_2026

A³M‑CIR: An Attention-Aware Adversarial Masking Based Self-Supervised Chemometric Framework for Robust Infrared Spectral Analysis Durgesh Ameta (21439753) Ajeet Kumar (141578) Tanishk Saini (24905864) Laxmidhar Behera (14341372) Space Science Neuroscience Pharmacology Biotechnology Chemical Sciences not elsewhere classified Biological Sciences not elsewhere classified Mental Health materials science applications global spectral context framework performs large emulate realistic perturbations dimensional vibrational signals aware adversarial masking varying noise conditions supervised chemometric framework noise ratio analyses based chemometric methods 3 </ sup limited labeled data supervised pretraining supervised baselines measurement noise limited supervision based self acquisition noise work establishes widely used validate robustness traditional preprocessing systematic signal scale self representative self often constrained manual annotation local continuity lightweight dual infrared spectroscopy improved performance extensive evaluations extensible foundation encoder captures driven pipelines dependent variability core component baseline drift Infrared (IR) spectroscopy provides rich molecular fingerprint information and is widely used for qualitative and quantitative chemical analysis across biomedical, environmental, and materials science applications. However, its practical utilization is often constrained by instrument-dependent variability, baseline drift, acquisition noise, and limited labeled data. While deep-learning-based chemometric methods have improved performance over traditional preprocessing-driven pipelines, their reliance on supervised learning limits robustness and transferability across heterogeneous measurement conditions. To address these challenges, we introduce A³M-CIR, an attention-aware adversarial masking-based self-supervised chemometric framework for robust infrared spectral analysis. The framework performs large-scale self-supervised pretraining on diverse unlabeled IR spectra to learn transferable representations without explicit calibration or manual annotation. Its core component, the Attention-Aware Adversarial Feature Masking Block (A³FMB), suppresses diagnostically salient absorption regions to generate informative contrastive views that emulate realistic perturbations, while a lightweight Dual-Mixing Attention (DMA) encoder captures both global spectral context and local continuity of one-dimensional vibrational signals. Extensive evaluations under linear-probing and fine-tuning protocols demonstrate consistent improvements over representative self-supervised baselines, and systematic signal-to-noise ratio analyses further validate robustness and transferability under varying noise conditions. By explicitly addressing spectral variability, measurement noise, and limited supervision as intrinsic challenges of infrared spectroscopy, this work establishes A³M-CIR as a practical and extensible foundation for data-driven infrared spectral analysis. 2026-09-14T00:00:00Z Dataset Dataset 10.1021/acs.analchem.6c01881.s001 https://figshare.com/articles/dataset/A_sup_3_sup_M_CIR_An_Attention-Aware_Adversarial_Masking_Based_Self-Supervised_Chemometric_Framework_for_Robust_Infrared_Spectral_Analysis/33763979 CC BY-NC 4.0