A Generative AI Approach for Reducing Skin Tone Bias in Skin Cancer Classification

Areez Muhammed Shabu; Mohammad Samar Ansari; Asra Aslam

cs.CVFeb 16, 2026

A Generative AI Approach for Reducing Skin Tone Bias in Skin Cancer Classification

Authors: Areez Muhammed Shabu, Mohammad Samar Ansari, Asra Aslam

Organizations: University of Sheffield, UK · University of Chester, UK

Abstract

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.

Figures & tables

Figure 1. Sample of a non-melanoma and melanoma image from ISIC dataset.

Figure 2. End-to-end pipeline. The Data Preparation branch (left) analyses the ISIC dataset for skin tone distribution and isolates the underrepresented dark-skin subset (FST V–VI, 1,407 images). The Generation Branch (right) fine-tunes a Stable Diffusion model via LoRA on this subset, generates 808 synthetic images, and validates them using statistical similarity metrics. In the Integration stage, original and synthetic data are merged and preprocessed. This stage also involved validation of augmented dataset with segmentation. The Classification Branch trains an EfficientNet-B0 and evaluates performance.

Figure 3. Generated Images of a non-melanoma and melanoma using proposed Gen AI-Diffusers based Pipeline

Skin Tone (FST)	Melanoma	Non-Mel.	Total
Brown (V)	8	47	55
Dark Brown/Black (VI)	215	1,137	1,352
Intermediate (III)	2,005	7,756	9,761
Light (II)	506	4,382	4,888
Tan (IV)	426	1,213	1,639
Uncertain	5	28	33

Table 1. Skin tone distribution in the ISIC dataset, classified by ITA-derived Fitzpatrick type and lesion category.

Figure 4. Distribution of melanoma and non-melanoma images across Fitzpatrick skin types (log scale). The shaded region highlights FST V and VI, which together represent under 8% of the dataset.

Figure 5. Training dynamics of the augmented EfficientNet-B0 model (a) loss, (b) accuracy, (c) validation AUC.

Class	Original	Synthetic	Augmented
Melanocytic	3,165 (223 dark)	144	3,309
Non-melanocytic	14,563 (1,184 dark)	664	15,227
Total	17,728	808	18,536

Table 2. Composition of the original and augmented datasets. Synthetic images target dark skin tones exclusively.

	Max-Flow	Orig	Gen-AI
Metric	(Non-ML)	Dataset	Dataset
	Algorithm	CNN	CNN
Mean IoU	0.75	0.82	0.85
Dice Coefficient	0.82	0.88	0.90
Precision	0.79	0.86	0.89
Recall (Sensitivity)	0.77	0.84	0.87

Table 3. Segmentation performance for vlidation of Gen AI dataset. Comparison of original of Non-ML and CNN based method and testing of Gen-AI synthetic dataset on best performing method (CNN). The Gen-AI CNN was trained on real + synthetic data; baseline CNN on real data. Bold indicates best result per metric.

Algorithm/Task	Mean Runtime
Segmentation (Augmented CNN)	130 ms/image (inference)
Synthetic Generation (Diffusers)	$\sim$ 4–6 sec/image
Classification (EffNet-B0)	60 ms/image

Table 4. Mean Runtimes on Stanage HPC

	Baseline Dataset		Gen AI based Dataset
Metric	Train	Val	Train	Val
Loss	0.33	0.47	0.15	0.23
Accuracy	87.65	85.45	94.00	92.14
AUC	0.988	0.9765	0.9725	0.9480
Precision	0.85	0.82	0.91	0.89
Recall	0.82	0.78	0.89	0.87

Table 5. Classification metrics for baseline (original data) and augmented (real + synthetic) models.

Explore similar work

Jul 14, 2026

Controllable Generation of Diverse Dermatological Imagery for Fair and Efficient Malignancy Classification

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.

Héctor Carrión, Narges Norouzi

University of California, Santa Cruz, USA · University of California, Berkeley, USA

Skin Lesion Classification Synthetic Data Augmentation Algorithmic Fairness Medical Image Classification Medical Image Synthesis Controllable Image Generation

Ask PDF

Mar 2, 2026

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

Skin tone classification is a fundamental task in computer vision, with applications in quantifying medical features, assessing demographic representativeness, and mitigating algorithmic bias. However, existing datasets either rely on a restrictive 6-tone Fitzpatrick scale, prohibit training through restrictive licenses, or lack representation of extreme skin tones, while current methods often fail to disentangle true skin tone from in-the-wild imaging conditions, degrading to random-classifier performance. We introduce skin tone in the wild (STW), the first open-access dataset annotated with the 10-tone Monk Skin Tone (MST) scale, containing over 42K images and specifically curated to include underrepresented tones. Alongside a rigorous evaluation protocol, we propose SkinToneNet, a ViT-based model that leverages global image context for robust in-the-wild generalization. Experiments show that SkinToneNet achieves 90% weighted off-by-one accuracy, substantially outperforming prior baselines. We further use SkinToneNet to audit widely used face datasets, including CelebA and VGGFace2, revealing substantial representation imbalances across MST tones.

Vitor Matias, Márcus Vinícius Lobo Costa, Tiago Novello +1

Instituto de Ciências Matemáticas e de Computação (ICMC), Universidade de São Paulo (USP), São Carlos, São Paulo, Brazil · Instituto de Matemática Pura e Aplica, Rio de Janeiro, Brazil

Classification Algorithmic Fairness Image Classification Fine-Grained Image Classification

Ask PDF

Jun 1, 2026

FairGen: Preference-Aligned Diffusion for Demographically Equitable Medical Image Synthesis

Medical imaging is central to modern diagnostics, and artificial intelligence (AI) systems are increasingly used to support image-based analysis by improving efficiency, accuracy, and access to care. However, inequities in healthcare access and differential disease prevalence create severe demographic imbalances in clinical image data. Such imbalances are compounded by the fact that diseases can manifest with distinct features across demographic groups, rendering certain phenotypic presentations naturally rare. AI models trained on such imbalanced data risk perpetuating diagnostic bias and widening healthcare disparities. Here we introduce FairGen, a fairness-aware diffusion framework that synthesizes demographically balanced medical images while preserving pathology-relevant visual features. By embedding physician-aligned preferences into the generation process, FairGen improves subgroup coverage during synthesis and downstream classification. Applied to dermatology, radiology, and neuroimaging benchmark tasks, FairGen achieves fairness improvements of 95.9% for skin images, 80.0% for chest radiography, and 35.2% for brain MRI, while maintaining competitive diagnostic accuracy relative to models trained on original clinical data. Clinician-facing expert review and external validation on independent cohorts further support that these gains extend beyond standard fidelity metrics and are not confined to the original in-distribution datasets.

Zhimin Li, Ruichen Zhang, Zhen Tan +3

University of Pittsburgh, Swanson School of Engineering, Pittsburgh, 15261, USA · 2The University of North Carolina at Chapel Hill, Department of Computer Science, Chapel Hill, 27599, USA · 3Arizona State University, School of Computing and Augmented Intelligence, Tempe, 85281, USA +1

Diffusion Model Alignment Synthetic Data Augmentation Algorithmic Fairness Diffusion-Based Image Generation Medical Image Synthesis

Ask PDF