Datasets:
Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -32,6 +32,13 @@ tags:
|
|
| 32 |
|
| 33 |
> **This card describes a [DetectionBench](https://github.com/dronefreak/DetectionBench) dataset adapter for LLVIP. It does NOT host or redistribute the dataset itself -- LLVIP has an explicit non-commercial license with no redistribution grant. See [Getting the Data](#getting-the-data) for the official download links.**
|
| 34 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 35 |
## Disclaimer
|
| 36 |
|
| 37 |
DetectionBench is **not** an official release of LLVIP and does not host any LLVIP images, annotations, or derived files anywhere -- not on Hugging Face, not in this repository.
|
|
@@ -46,7 +53,7 @@ What this repository provides instead:
|
|
| 46 |
|
| 47 |
---
|
| 48 |
|
| 49 |
-
# Dataset
|
| 50 |
|
| 51 |
LLVIP is a registered visible/infrared image-pair dataset for pedestrian detection in low-light conditions: 30,976 image pairs (paired visible + infrared frames, pixel-aligned) captured at night, annotated with pedestrian bounding boxes. Its central finding is that infrared imagery is dramatically more useful than visible-light imagery for detection under these conditions -- the paired visible frames are frequently near-black.
|
| 52 |
|
|
|
|
| 32 |
|
| 33 |
> **This card describes a [DetectionBench](https://github.com/dronefreak/DetectionBench) dataset adapter for LLVIP. It does NOT host or redistribute the dataset itself -- LLVIP has an explicit non-commercial license with no redistribution grant. See [Getting the Data](#getting-the-data) for the official download links.**
|
| 34 |
|
| 35 |
+
## Dataset Description
|
| 36 |
+
|
| 37 |
+
- **Homepage:** https://bupt-ai-cz.github.io/LLVIP/
|
| 38 |
+
- **Repository:** https://github.com/bupt-ai-cz/LLVIP
|
| 39 |
+
- **Paper:** https://arxiv.org/abs/2108.10831
|
| 40 |
+
- **Point of Contact:** czhu@bupt.edu.cn
|
| 41 |
+
|
| 42 |
## Disclaimer
|
| 43 |
|
| 44 |
DetectionBench is **not** an official release of LLVIP and does not host any LLVIP images, annotations, or derived files anywhere -- not on Hugging Face, not in this repository.
|
|
|
|
| 53 |
|
| 54 |
---
|
| 55 |
|
| 56 |
+
# Dataset Overview
|
| 57 |
|
| 58 |
LLVIP is a registered visible/infrared image-pair dataset for pedestrian detection in low-light conditions: 30,976 image pairs (paired visible + infrared frames, pixel-aligned) captured at night, annotated with pedestrian bounding boxes. Its central finding is that infrared imagery is dramatically more useful than visible-light imagery for detection under these conditions -- the paired visible frames are frequently near-black.
|
| 59 |
|