RASMD: RGB And SWIR Multispectral Driving Dataset for Robust Perception in Adverse Conditions

Current autonomous driving algorithms heavily rely on the visible spectrum, which is prone to performance degradation in adverse conditions like fog, rain, snow, glare, and high contrast. Although other spectral bands like near-infrared (NIR) and long-wave infrared (LWIR) can enhance vision perception in such situations, they have limitations and lack large-scale datasets and benchmarks. Short-wave infrared (SWIR) imaging offers several advantages over NIR and LWIR. However, no publicly available large-scale datasets currently incorporate SWIR data for autonomous driving. To address this gap, we introduce the RGB and SWIR Multispectral Driving (RASMD) dataset, which comprises 100,000 synchronized and spatially aligned RGB-SWIR image pairs collected across diverse locations, lighting, and weather conditions. In addition, we provide a subset for RGB-SWIR translation and object detection annotations for a subset of challenging traffic scenarios to demonstrate the utility of SWIR imaging through experiments on both object detection and RGB-to-SWIR image translation. Our experiments show that combining RGB and SWIR data in an ensemble framework significantly improves detection accuracy compared to RGB-only approaches, particularly in conditions where visible-spectrum sensors struggle. We anticipate that the RASMD dataset will advance research in multispectral imaging for autonomous driving and robust perception systems.

Dataset Details

Folders and their content:

• RASMD_aligned: RGB-SWIR pair aligned images
• RASMD_Source_#: RGB-SWIR pair source images
• RASMD_detection: RGB-SWIR pair object detection benchmark images
• RASMD_detection_annotation: cocoformat object detection annotations
• RASMD_translation: image-to-image translation benchmark with zeroshot
• RASMD_metadata: RASMD_aligned, RASMD_source_# metadata json file. Included RGB-SWIR aligned, source pair, location(urban, suburban), weather(sunny, cloudy, rainy, snowy).

Dataset Description

• Curated by: Seamless Trans-X Lab of Yonsei University
• License: CC BY-NC-SA 4.0

Dataset Sources

• arXiv Paper: https://arxiv.org/abs/2504.07603
• Webpage: yonsei-stl.github.io/RASMD/

Dataset Structure

-RASMD_aligned, RASMD_source

├── RASMD_DATASET (RASMD_aligned)
│   ├── 1_sunny
│   │   ├── data_1_urban
│   │   │   └── croped_data
│   │   │       ├── rgb
│   │   │       │   └── ...
│   │   │       └── swir
│   │   │           └── ...
│   │   └── data_2_urban
│   │       ...       
│   └── 2_sunny
│       ... 
│
├── RASMD_source
│   ├── 1_sunny
│   │   ├── data_1_urban
│   │   │   └── source_images
│   │   │       ├── rgb
│   │   │       │   └── ...
│   │   │       └── swir
│   │   │           └── ...
│   │   └── data_2_urban
│   │       ... 
│   └── 2_sunny
│       ... 

-RASMD_detection

├── RASMD_detection
│   ├── train
│   │   ├── RGB
│   │   │   └── ...
│   │   └── SWIR
│   │       └── ...     
│   ├── test (Domain only test)
│   │   ├── RGB
│   │   │   └── ...
│   │   └── SWIR
│   │       └── ... 
│   ├── test_mgerged (Cross domain test)
│   │   ├── rgb
│   │   │   └── ...
│   │   └── swir
│   │       └── ...

-RASMD_detection_annotation

├── RASMD_detection_annotation
│   ├── train_rgb_align.json   (->train/RGB)
│   ├── train_swir_align.json  (->train/SWIR)
│   ├── test_rgb_align.json    (->test/RGB)
│   ├── test_swir_align.json   (->test/SWIR)
│   ├── test_merged_rgb.json   (->test_merged/rgb) 
│   └── test_merged_swir.json  (->test_merged/swir)

-RASMD_translation

├── RASMD_translation 
│   ├── train
│   │   ├── RGB
│   │   │   └── ...
│   │   └── SWIR
│   │       └── ...     
│   ├── test 
│   │   ├── RGB
│   │   │   └── ...
│   │   └── SWIR
│   │       └── ... 
│   ├── Zeroshot
│   │   ├── train
│   │   │   ├── RGB
│   │   │   │   └── ...
│   │   │   └── SWIR
│   │   │       └── ...     
│   │   ├── test 
│   │   │   ├── RGB
│   │   │   │   └── ...
│   │   │   └── SWIR
│   │   │       └── ...

-RASMD_metadata.json

{
    "data": [
        {
            "unique_id": 
            "align_pair": {
                "rgb": {
                    "path": 
                    "width":
                    "height": 
                },
                "swir": {
                    "path": 
                    "width": 
                    "height": 
                }
            },
            "source_pair": {
                "rgb": {
                    "path": 
                    "width": 
                    "height": 
                },
                "swir": {
                    "path": 
                    "width": 
                    "height": 
                }
            },
            "weather": 
            "location": 
        },
        ...

Annotations

Object detection annotation distribution

Citation

BibTeX:

@misc{jin2025rasmdrgbswirmultispectral,
      title={RASMD: RGB And SWIR Multispectral Driving Dataset for Robust Perception in Adverse Conditions}, 
      author={Youngwan Jin and Michal Kovac and Yagiz Nalcakan and Hyeongjin Ju and Hanbin Song and Sanghyeop Yeo and Shiho Kim},
      year={2025},
      eprint={2504.07603},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2504.07603}, 
}

Contact

• Youngwan Jin ([email protected])
• Shiho Kim ([email protected])