Applied Machine Learning in the field of Facial Aesthetics to forecast variances in subject trajectories before and after synthetic facial treatments (using GANs). This pivotal step enables patients to visualize their potential appearance at various stages of their facial treatments prior to commencing the actual procedure.

• CNN - Severity of Crows Feet

  Built a custom GPU-based CNN Pipeline which included retraining an InceptionV3 model to detect Wrinkles and Fine Lines in targeted areas of the face (Crows Feet).

• Age Prediction & GANS

  Designed and implemented a novel evaluation metric, known as the Age Prediction Metric, to accurately forecast the ages of subjects, both placebo and those subjected to synthetic facial treatments using GANs.

• StyleGANs - undersampled groups

  Leveraged StyleGANS to generate synthetic images of undersampled groups, addressing the challenge of limited data availability in these specific demographics.