SMAP
Datasets
All datasets matching “SMAP”SMAP_L3_SM_PNASA Earthdata - Soil data (SMAP_L3_SM_P)
Date range:
2024-03-01 ~ 2025-09-05 (.h5 files)
2022-01-01 ~ 2025-09-05 (.parquet files)
Content:
.h5 files (~32MB per file, raw, includes the daily world data)
.parquet files (<600 KB per file)
smap_hungary_YYYY_cells.parquet → raw per-pass per-cell values (full granularity).
smap_hungary_YYYY_cells_index.parquet → manifest/metadata describing coverage + kept cells per pass.
smap_hungary_YYYY_cells_agg_7d.parquet → 7-day aggregated per-cell values… See the full description on the dataset page: https://huggingface.co/datasets/jaelin215/SMAP_L3_SM_P.SMAP
SMAP (TsFile)
Apache TsFile version of the SMAP anomaly-detection subset of
thuml/Time-Series-Library.
Overview
Soil Moisture Active Passive satellite telemetry (25 channels) for anomaly detection.
Train: 135,183 rows (all normal).
Test: 427,617 rows (with per-timestep 0/1 anomaly labels).
Channels: 25.
The train and test segments are stored as two separate TsFiles
(SMAP_train.tsfile / SMAP_test.tsfile), preserving the original split.
Schema (TsFile… See the full description on the dataset page: https://huggingface.co/datasets/THULab/SMAP.smapper-light
📂 SMapper-light Dataset
SMapper-light is a publicly available multimodal dataset collected using the SMapper platform, an open-hardware, multi-sensor device designed for SLAM (Simultaneous Localization and Mapping) research.
The dataset provides synchronized LiDAR, multi-camera, and IMU measurements, enabling benchmarking of visual, LiDAR, and visual–inertial SLAM methods.
🚀 Applications
SMapper-light can be used for:
Benchmarking LiDAR SLAM frameworks (e.g.… See the full description on the dataset page: https://huggingface.co/datasets/snt-arg/smapper-light.Q-Spatial-Bench-sMAPE-Comparison
Q-Spatial-Bench-sMAPE-Comparison
Recording sMAPE for models evaluated on Q-Spatial-Bench across model families, sizes, finetuning, prompting, post-training optimization strategies.
log_gene_smapledpsychopathology-labelled-reddit