Demographically oriented facial attribute recognition for skincare and grooming: dataset, benchmark, and privacy-aware reporting pipeline

Facial attribute recognition plays a crucial role in applications ranging from human–computer interaction to personalized digital health. However, the effectiveness of existing systems is often limited by demographic bias in training data and the absence of domain-specific annotations, particularly for nuanced tasks such as skincare and grooming analysis. Large-scale datasets like CelebA are predominantly Western-centric and lack critical attributes including Oily Skin, Wrinkles, and grooming-related characteristics. To address these limitations, we introduce the Indian Skincare and Grooming Dataset (ISGD), a manually curated dataset comprising 30,141 facial images from the Indian subcontinent, annotated across 33 fine-grained binary attributes specifically designed for skincare and grooming analysis. Building upon ISGD, we propose AKRTI , a privacy-conscious inference pipeline that decouples visual processing from report generation. The system employs a ConvNeXt-Tiny backbone for multi-label facial attribute prediction. Importantly, only the predicted binary attribute vector—never the raw facial image—is passed to a large language model (LLM) to generate a personalized, human-readable skincare and grooming report, thereby limiting exposure of raw biometric data to the report-generation stage. We position ISGD as the primary contribution of this work—a demographically oriented benchmark for fine-grained facial attribute recognition—and AKRTI as one privacy-conscious downstream application that demonstrates the dataset’s utility. Experimental results demonstrate that models trained on ISGD significantly outperform those trained on a size-matched subset of CelebA, achieving 94.26% overall accuracy and an F1-score of 0.8851. Furthermore, per-attribute evaluation indicates more consistent and reliable predictions for skincare-critical features such as beard presence, skin condition, and wrinkles. By introducing a demographically oriented dataset alongside a privacy-aware reporting framework, this work establishes a foundation for equitable and practical AI-driven facial analysis systems in personalized wellness applications. AKRTI is presented as a wellness-oriented exploratory tool and is not intended for clinical or diagnostic use. To facilitate reproducible research, we publicly release the dataset, annotation guidelines, training and evaluation scripts, model checkpoints, and the AKRTI inference prompt. Source code is available at our GitHub repository: https://github.com/HimalRana2610/ISGD . The work is archived at Zenodo (DOI: 10.5281/zenodo.18837811 ).