tinyvit

Lightweight Vision Transformer models for multiclass skin-lesion classification using imbalance-aware learning, Grad-CAM visualization, and efficiency analysis.

computer-vision deep-learning pytorch medical-imaging class-imbalance gradcam skin-lesion-classification ham10000 vision-transformer mobilevit tinyvit

• Updated Aug 5, 2026
• Jupyter Notebook

Two-stage Landsat-9 TIR pipeline for 200m→100m super-resolution and RGB-like colorization using CNN/U-Net models, raw sensor preprocessing, metric evaluation, and deployable inference dashboard.

dashboard deep-learning image-processing cnn pytorch remote-sensing gan image-colorization satellite-imagery unet pix2pix super-resolution earth-observation model-evaluation thermal-infrared tir landsat-9 tinyvit

• Updated Jul 5, 2026
• Jupyter Notebook