The study audits four image aesthetic scorers—LAION-Aesthetics, PickScore, ImageReward, and HPSv2—using pixel‑level interventions on skin tone and body type in both synthetic and real images. It finds that most scorers exhibit a fidelity preference: unaltered images receive the highest scores, while perturbations in either direction are penalized in an inverted‑U pattern, and this effect is largely independent of the skin operator. Synthetic‑only audits are misleading, as the apparent preference for darker skin in synthetic faces reverses or weakens when evaluated on real faces, and cross‑scorer results vary widely, underscoring the need for real‑data, within‑image causal isolation to accurately assess demographic bias.
arXiv:2407. 13922v3 Announce Type: replace-cross Abstract: Face recognition (FR) systems are widely deployed in critical applications, making their reliability and robustness across diverse populations and conditions essential.
arXiv:2606. 20094v1 Announce Type: cross Abstract: Makeup transfer models enable fun augmented reality (AR) experiences as well as virtual try-on (VTO) for online makeup shopping.
arXiv:2609.36400v1 Announce Type: cross Abstract: Deep learning classifiers for dermoscopic skin lesions often reach high in-distribution accuracy while quietly relying on spurious background cues su...
The study examines why dermatology AI models, largely trained on light‑skinned, cancer‑focused images, perform poorly when applied to diverse patient populations. By comparing a cancer‑trained baseline, two dermatology foundation models, and a general‑purpose vision model on tone‑stratified and disease‑shifted datasets, the authors find that disease‑distribution shift, rather than skin‑tone underrepresentation, is the primary cause of generalization failure. Representation analysis shows that cancer‑specialized features lack transferable structure, while dermatology‑pretrained features maintain stronger clustering, and lightweight adaptation with about ten labeled examples per category can recover most performance.
Fairness evaluation in computer vision commonly relies on aggregate accuracy and demographic subgroup analysis. However, visual models are also sensitive to contextual factors such as illumination, blur, image quality, facial accessories, and appearance attributes.
arXiv:2605. 20731v2 Announce Type: replace-cross Abstract: Text-to-image models now generate graphic design at production scale, yet their supervision still comes primarily from photo-style preference datasets with a single overall verdict per comparison.
The paper introduces a geometry‑driven shadow harmonisation pipeline for composited faces that preserves albedo by applying a per‑pixel gain field derived from a rasterised 3D face proxy. The method estimates key‑light direction from host cues, computes surface normals and cast shadows, and applies a controlled, monotonic darkening operation that never alters chromaticity or identity. Experiments on analytic face heightfields show that the default parameters modify over half the face pixels while maintaining hue invariance and avoiding repainting the donor face.
arXiv:2608. 01301v3 Announce Type: replace-cross Abstract: Infrared-visible image fusion (IVIF) has no ideal fused reference, so algorithms are ranked by scalar objective metrics that formalize proxies for information transfer, structure, or source similarity.
arXiv:2606. 18451v1 Announce Type: new Abstract: Single-image-to-3D generators are improving quickly, but there is no agreed, human-free way to tell whether one generated mesh is better than another.
arXiv:2603. 10834v3 Announce Type: replace-cross Abstract: Understanding how neural networks rely on visual cues offers a human-interpretable view of their internal decision processes.
arXiv:2606. 20364v1 Announce Type: new Abstract: A companion study established a de-biased, cross-model VLM-as-3D-judge that reliably ranks single-image-to-3D mesh quality where cheap geometry and CLIP proxies fall short.
arXiv:2607. 21434v1 Announce Type: cross Abstract: Video face swapping has no natural paired supervision: no real footage exists of one person's face performing another person's video.