Pixel Art Normal Maps is a synthetic dataset of image pairs for training and evaluating normal-map generation models for 2D pixel art. It accompanies a master's thesis on normal-map generation for pixelated images. The dataset was created by DmytroKhitro and is hosted on Hugging Face.
Use Cases
- Training normal-map generation models based on synthetic pixel-art sprite to normal-map pairs.
- Evaluating model performance for 2D pixel-art enhancement based on the described image pairs.
- Researching computer graphics techniques for pixelated images based on the dataset's synthetic nature.
Strengths
- Dataset is specifically designed for training and evaluating normal-map generation models.
- Dataset accompanies a master's thesis, suggesting a research-oriented purpose.
Limitations
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count is unknown, which may limit suitability assessment.
- Last updated 2026-06-05 03:42:56; freshness should be verified.
Provenance
- Source
- DmytroKhitro
- Collection Method
- Synthetic generation (pixel-art sprite → camera-space normal map).