SmellNet is a comparatively large dataset for sensor-based machine olfaction. It contains real-world smell measurements collected from a compact array of low-cost metal-oxide gas sensors. The dataset was created by DeweiFeng and was last updated on April 13, 2026.
Use Cases
- Single-substance recognition based on multichannel sensor time series.
- Mixture distribution prediction based on odor mixtures.
- Cross-modal learning based on paired GC-MS-derived chemistry priors.
- Research on temporal modeling for smell sensing based on sensor time series data.
Strengths
- Described as a comparatively large dataset for its domain.
- Contains real-world measurements from a compact array of low-cost sensors.
- Supports multiple research tasks including single-substance recognition and mixture prediction.
Limitations
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count and overall data scale are unknown, which may limit suitability assessment.
- Description metadata is limited; actual data quality requires manual inspection after download.
Provenance
- Source
- huggingface, author DeweiFeng
- Collection Method
- Real-world smell measurements collected from a compact array of low-cost metal-oxide gas sensors.
- Freshness
- Last updated 2026-04-13 08:13:14.