Skin Tone in the Wild: A Dataset and Benchmark for 10-tone MST Classification

Skin cancer is one of the most common cancers worldwide and early detection is critical for effective treatment. However, current AI diagnostic tools are often trained on datasets dominated by lighter skin tones, leading to reduced accuracy and fairness for people with darker skin. The International Skin Imaging Collaboration (ISIC) dataset, one of the most widely used benchmarks, contains over 70% light skin images while dark skins fewer than 8%. This imbalance poses a significant barrier to equitable healthcare delivery and highlights the urgent need for methods that address demographic diversity in medical imaging. This paper addresses this challenge of skin tone imbalance in automated skin cancer detection using dermoscopic images. To overcome this, we present a generative augmentation pipeline that fine-tunes a pre-trained Stable Diffusion model using Low-Rank Adaptation (LoRA) on the image dark-skin subset of the ISIC dataset and generates synthetic dermoscopic images conditioned on lesion type and skin tone. In this study, we investigated the utility of these images on two downstream tasks: lesion segmentation and binary classification. For segmentation, models trained on the augmented dataset and evaluated on held-out real images show consistent improvements in IoU, Dice coefficient, and boundary accuracy. These evalutions provides the verification of Generated dataset. For classification, an EfficientNet-B0 model trained on the augmented dataset achieved 92.14% accuracy. This paper demonstrates that synthetic data augmentation with Generative AI integration can substantially reduce bias with increase fairness in conventional dermatological diagnostics and open challenges for future directions.

Skin diseases represent a major global public health burden, yet machine learning tools developed to assist in their diagnosis suffer from two critical limitations: reliance on only one modality for diagnosis and systematic performance disparities across skin tones. While existing approaches address each challenge separately, this work proposes a modality-invariant framework with fair representation (MIFR) for skin disease classification. The architecture pairs clinical photographs with dermoscopic images using ViT-based encoders, projecting each input into a high-dimensional embedding space via modality-specific projection heads. The resulting model is trained with a five-component multi-objective loss including weighted cross-entropy for classification, confusion and skin-type classification losses for fairness, per-modality supervised contrastive loss for class alignment, and a modality-invariance loss for clinical and dermoscopic modality alignment. Experiments on the HIBA+Derm7pt paired dataset and the external PAD-UFES-20 and ISIC 2019 datasets showed that modality-invariant representation learning provides competitive predictive performance compare to relevant baseline models and competitive fairness on the internal dataset. t-SNE visualizations confirmed that clinical and dermoscopic embeddings of the same disease are geometrically aligned, validating the joint objectives.

Accurate dermatological diagnosis naturally necessitates equitable performance across diverse populations, yet a systematic lack of expertly annotated images, especially for underrepresented skin tones and rare diseases, impedes progress toward measurably fair methods. We introduce cgDDI (Controllable Generation of Diverse Dermatological Imagery), a hybrid framework that (1) synthesizes realistic healthy skin samples without disturbing other input properties, (2) maps single-sample rare lesions onto novel skin-tones and locations non-parametrically, and (3) allows for efficient parametric generation with as few as 10 training samples. The framework supports both human and automated segmentation masking, enabling scalability to datasets without pre-made lesion masks. We grow a 656-image dataset by more than 400x and validate across two datasets: biopsy-confirmed Diverse Dermatology Images (DDI) and expert-verified Fitzpatrick17k (F17k). On the DDI benchmark, we achieve malignancy classification accuracy of 86.4% under synthetic-only training and 90.9% state-of-the-art performance with real data fine-tuning, alongside leading fairness metrics. Cross-dataset experiments show +13.9% accuracy improvements on unseen F17k data despite minimal disease overlap. We openly release 266k+ synthetic images, code, and generative models to further support fairness research at https://github.com/hectorcarrion/ControllableGenDDI.