How it works
Hair Volume first isolates uncovered hair with the same six-class, face-anchored portrait mask used by Hair Color. It then expands the existing hair texture by a small, intensity-controlled amount—up to 14% horizontally and 8% vertically—and permits new pixels only where the model independently identifies surrounding background. Face skin, body skin, clothing, and accessories are forbidden destinations. The tool never generates a new face or hairstyle, returns a lossless PNG, and copies every pixel outside the expanded hair silhouette unchanged. If the photo isn't clear enough for this, the same change is made to the whole photo with AI instead.

























