Datasets › LLVIP

LLVIP (A Visible-infrared Paired Dataset for Low-light Vision)

Introduced by Xinyu Jia et al. in LLVIP: A Visible-infrared Paired Dataset for Low-light Vision24 Aug 2021 archive 2025-07-28

Visible-infrared Paired Dataset for Low-light Vision
30976 images (15488 pairs)
24 dark scenes, 2 daytime scenes
Support for image-to-image translation (visible to infrared, or infrared to visible), visible and infrared image fusion, low-light pedestrian detection, and infrared pedestrian detection
(The original image and video pairs (before registration) of LLVIP are also released!)

Homepage Benchmarks Papers JSON

Benchmarks archive 2025-07-28

All 6 leaderboards whose dataset resolves to this page shown (sort by any header). "First row" is the archive's own first row at snapshot, in the archive's row order; nothing here re-ranks and metric direction is not asserted.

		First row (archive order)	Paper	Code
Pedestrian Detection	LLVIP	MMPedestron AP 0.726	When Pedestrian Detection Meets Multi-Modal Learning:...	BubblyYi/MMPedestron	15	Compare
Image-to-Image Translation	LLVIP	pyramidpix2pix PSNR 12.191	BCI: Breast Cancer Immunohistochemical Image Generation...	bupt-ai-cz/BCI	4	Compare
Thermal Infrared Pedestrian Detection	LLVIP	YoloV5 AP 0.670	LLVIP: A Visible-infrared Paired Dataset for Low-light Vision	bupt-ai-cz/LLVIP	2	Compare
Image Generation	LLVIP	pix2pix PSNR 10.769	LLVIP: A Visible-infrared Paired Dataset for Low-light Vision	bupt-ai-cz/LLVIP	1	Compare
Multispectral Object Detection	LLVIP	CFT mAP50 97.5	Cross-Modality Fusion Transformer for Multispectral...	docf/multispectral-object-detection	1	Compare
Object Detection	LLVIP	MiPa AP 0.665	MiPa: Mixed Patch Infrared-Visible Modality Agnostic...	heitorrapela/mipa	1	Compare

Papers archive 2025-07-28

12 shown of 12 papers with a leaderboard row on this dataset's benchmarks, newest first. The archive's own "papers using this dataset" list was never published, so this is the benchmark-backed subset; the archive's count for this dataset is 116. The Syntology column is from Syntology's graph (read 2026-09-24), stated per sample; it is not part of any archive number.

			Date	Samples run Syntology
CAFF-DINO: Multi-spectral object detection transformers with cross-attention features fusion	0	1	27 Sep 2024	not harvested
When Pedestrian Detection Meets Multi-Modal Learning: Generalist Model and Benchmark Dataset	1	1	14 Jul 2024	not harvested
MiPa: Mixed Patch Infrared-Visible Modality Agnostic Object Detection	1	2	29 Apr 2024	not harvested
UniRGB-IR: A Unified Framework for RGB-Infrared Semantic Tasks via Adapter Tuning	1	1	26 Apr 2024	not harvested
INSANet: INtra-INter Spectral Attention Network for Effective Feature Fusion of Multispectral Pedestrian Detection	1	1	10 Feb 2024	not harvested
Removal then Selection: A Coarse-to-Fine Fusion Perspective for RGB-Infrared Object Detection	1	1	19 Jan 2024	not harvested
Multimodal Object Detection by Channel Switching and Spatial Attention	0	4	18 Jun 2023	not harvested
BCI: Breast Cancer Immunohistochemical Image Generation through Pyramid Pix2pix	1	3	25 Apr 2022	not harvested
CMX: Cross-Modal Fusion for RGB-X Semantic Segmentation with Transformers	1	1	9 Mar 2022	ran 2 of 2 samples (0 unverified)
Cross-Modality Fusion Transformer for Multispectral Object Detection	1	2	30 Oct 2021	not harvested
LLVIP: A Visible-infrared Paired Dataset for Low-light Vision	1	6	24 Aug 2021	not harvested
Multimodal Object Detection via Probabilistic Ensembling	3	1	7 Apr 2021	ran 8 of 11 samples (3 unverified; 8 pointer-only for licence)

Dataset loaders archive 2025-07-28

bupt-ai-cz/LLVIPtfpytorch

1 loader as listed in the archive; links are outbound and not re-checked here.

Tasks archive 2025-07-28

Object Detection Image Generation Image-to-Image Translation Image Enhancement Low-Light Image Enhancement Pedestrian Detection Image Registration Multispectral Object Detection Zero-shot Classification (unified classes)Thermal Infrared Pedestrian Detection Infrared And Visible Image Fusion Low-light Pedestrian Detection

License archive 2025-07-28

This LLVIP Dataset is made freely available to academic and non-academic entities for non-commercial purposes such as academic research, teaching, scientific publications, or personal experimentation. Permission is granted to use the data given that you agree to our license terms.

Modalities archive 2025-07-28

Images

Languages archive 2025-07-28

English Chinese

Variants archive 2025-07-28

LLVIP

1 variant name, as the archive lists them.

At a glance archive 2025-07-28

Introduced: 24 Aug 2021
Papers: 116
Benchmarks: 6
With a result: 12
Licence: This LLVIP Dataset is made freely available to academic and non-academic entities for non-commercial purposes such as academic research, teaching, scientific publications, or personal experimentation. Permission is granted to use the data given that you agree to our license terms.
Modalities: Images
Languages: English, Chinese
Loaders: 1
Homepage: bupt-ai-cz.github.io/LLVIP/

Samples run Syntology

Of the 12 papers with a result here, 2 have harvested code samples in Syntology's graph. Summed over those: Syntology ran 10 of 13 samples; 3 unverified; 8 pointer-only for licence. 0 of the 2 have no sample that ran. Each row in the Papers table states its own "ran N of M samples" with the unverified count; a sample that ran executed on a synthesized fixture, which is not a correctness claim about any result. Read from the graph 2026-09-24.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections