NTUT 4K Drone Photo - Dataset Ninja

Introduction #

Kuan-Ting (K. T.) Lai

In the NTUT 4K Drone Photo Dataset for Human Detection authors furnish 4K photos extracted from drone videos captured in Taiwan. Authors claim, that contemporary drones are outfitted with 4K video cameras, and the heightened resolution of the images facilitates modern object detectors in discerning smaller objects. Despite this capability, many drone image datasets typically offer only downscaled images. The dataset is curated by the AIoT Lab at the National Taiwan University of Technology (NTUT).

Sometimes humans may be blocked. In this case, the following convention is proposed:

block25 (Object is blocked 25~50%)
block50 (Object is blocked 50~75%)
block75 (Object is blocked 75~90%)

Also, the dataset propose recognizable objects for the purpose of identification task.

Expand

Homepage

Blog Post

Competition

Summary #

NTUT 4K Drone Photo Dataset for Human Detection is a dataset for object detection and identification tasks. Possible applications of the dataset could be in the drone inspection domain.

The dataset consists of 4095 images with 50553 labeled objects belonging to 12 different classes including walk, recognizable, stand, and other: riding, sit, block50, block25, push, block75, watchphone, baseball, and soccer.

Images in the NTUT 4K Drone Photo dataset have bounding box annotations. There are 26 (1% of the total) unlabeled images (i.e. without annotations). There are 2 splits in the dataset: train (2156 images) and test (1939 images). Additionally, every recognizable class has its own id tag. The dataset was released in 2022 by the Taipei Tech AIoT Lab, Taiwan.

Explore #

NTUT 4K Drone Photo dataset has 4095 images. Click on one of the examples below or open "Explore" tool anytime you need to view dataset images with annotations. This tool has extended visualization capabilities like zoom, translation, objects table, custom filters and more. Hover the mouse over the images to hide or show annotations.

Sample annotation mask from NTUT 4K Drone Photo

👀

Have a look at 4095 images

Because of dataset's license preview is limited to 12 images

View images along with annotations and tags, search and filter by various parameters

Class balance #

There are 12 annotation classes in the dataset. Find the general statistics and balances for every class in the table below. Click any row to preview images that have labels of the selected class. Sort by column to find the most rare or prevalent classes.

Rows 1-10 of 12

Class ㅤ	Images ㅤ	Objects ㅤ	Count on image average	Area on image average
walk➔ rectangle	2486	9783	3.94	0.34%
recognizable➔ rectangle	1912	32388	16.94	2.32%
stand➔ rectangle	1291	3801	2.94	0.27%
riding➔ rectangle	681	2384	3.5	0.61%
sit➔ rectangle	428	905	2.11	0.17%
block50➔ rectangle	277	323	1.17	0.1%
block25➔ rectangle	241	282	1.17	0.14%
push➔ rectangle	176	177	1.01	0.13%
block75➔ rectangle	173	225	1.3	0.12%
watchphone➔ rectangle	147	167	1.14	0.12%

Co-occurrence matrix #

Co-occurrence matrix is an extremely valuable tool that shows you the images for every pair of classes: how many images have objects of both classes at the same time. If you click any cell, you will see those images. We added the tooltip with an explanation for every cell for your convenience, just hover the mouse over a cell to preview the description.

Images #

Explore every single image in the dataset with respect to the number of annotations of each class it has. Click a row to preview selected image. Sort by any column to find anomalies and edge cases. Use horizontal scroll if the table has many columns for a large number of classes in the dataset.

Object distribution #

Interactive heatmap chart for every class with object distribution shows how many images are in the dataset with a certain number of objects of a specific class. Users can click cell and see the list of all corresponding images.

Class sizes #

The table below gives various size properties of objects for every class. Click a row to see the image with annotations of the selected class. Sort columns to find classes with the smallest or largest objects or understand the size differences between classes.

Rows 1-10 of 11

Class	Object count	Avg area	Max area	Min area	Min height	Min height	Max height	Max height	Avg height	Avg height	Min width	Min width	Max width	Max width
recognizable rectangle	32388	0.15%	25.88%	0.01%	20px	0.93%	2160px	100%	145px	6.72%	12px	0.31%	996px	25.94%
walk rectangle	9783	0.09%	1.99%	0.01%	22px	1.02%	571px	26.44%	100px	4.64%	17px	0.44%	309px	8.05%
stand rectangle	3801	0.09%	3.83%	0.01%	28px	1.3%	569px	26.34%	93px	4.3%	22px	0.57%	649px	16.9%
riding rectangle	2384	0.18%	0.39%	0.01%	19px	0.88%	230px	10.65%	140px	6.48%	33px	0.86%	230px	5.99%
sit rectangle	905	0.08%	1.88%	0.02%	32px	1.48%	373px	17.27%	76px	3.53%	30px	0.78%	418px	10.89%
block50 rectangle	323	0.08%	1.14%	0.01%	22px	1.02%	404px	18.7%	86px	3.99%	25px	0.65%	234px	6.09%
block25 rectangle	282	0.12%	3.46%	0.01%	39px	1.81%	569px	26.34%	103px	4.78%	23px	0.6%	530px	13.8%
block75 rectangle	225	0.09%	0.43%	0.01%	19px	0.88%	197px	9.12%	87px	4.01%	20px	0.52%	246px	6.41%
push rectangle	177	0.12%	0.3%	0.03%	61px	2.82%	143px	6.62%	109px	5.02%	39px	1.02%	173px	4.51%
watchphone rectangle	167	0.11%	0.22%	0.05%	46px	2.13%	158px	7.31%	92px	4.26%	59px	1.54%	136px	3.54%

Spatial Heatmap #

The heatmaps below give the spatial distributions of all objects for every class. These visualizations provide insights into the most probable and rare object locations on the image. It helps analyze objects' placements in a dataset.

Objects #

Table contains all 50553 objects. Click a row to preview an image with annotations, and use search or pagination to navigate. Sort columns to find outliers in the dataset.

Rows 1-10 of 50553

Object ID ㅤ	Class ㅤ	Image name click row to open	Image size height x width	Height ㅤ	Height ㅤ	Width ㅤ	Width ㅤ	Area ㅤ
1➔	recognizable rectangle	Drone_049.mp4_t-26.166667.jpg	2160 x 3840	188px	8.7%	104px	2.71%	0.24%
2➔	recognizable rectangle	Drone_049.mp4_t-26.166667.jpg	2160 x 3840	168px	7.78%	58px	1.51%	0.12%
3➔	recognizable rectangle	Drone_049.mp4_t-26.166667.jpg	2160 x 3840	163px	7.55%	59px	1.54%	0.12%
4➔	recognizable rectangle	Drone_049.mp4_t-26.166667.jpg	2160 x 3840	156px	7.22%	70px	1.82%	0.13%
5➔	recognizable rectangle	Drone_049.mp4_t-26.166667.jpg	2160 x 3840	116px	5.37%	54px	1.41%	0.08%
6➔	recognizable rectangle	Drone_049.mp4_t-26.166667.jpg	2160 x 3840	169px	7.82%	100px	2.6%	0.2%
7➔	recognizable rectangle	Drone_049.mp4_t-26.166667.jpg	2160 x 3840	139px	6.44%	30px	0.78%	0.05%
8➔	recognizable rectangle	Drone_049.mp4_t-26.166667.jpg	2160 x 3840	88px	4.07%	52px	1.35%	0.06%
9➔	recognizable rectangle	Drone_049.mp4_t-26.166667.jpg	2160 x 3840	58px	2.69%	22px	0.57%	0.02%
10➔	recognizable rectangle	Drone_049.mp4_t-26.166667.jpg	2160 x 3840	144px	6.67%	62px	1.61%	0.11%

License #

License is unknown for the NTUT 4K Drone Photo Dataset for Human Detection dataset.

Source

Citation #

If you make use of the NTUT 4K Drone Photo data, please cite the following reference:

@dataset{NTUT 4K Drone Photo,
  author={Kuan-Ting (K. T.) Lai},
  title={NTUT 4K Drone Photo Dataset for Human Detection},
  year={2022},
  url={https://www.kaggle.com/datasets/kuantinglai/ntut-4k-drone-photo-dataset-for-human-detection/data}
}

Source

If you are happy with Dataset Ninja and use provided visualizations and tools in your work, please cite us:

@misc{ visualization-tools-for-ntut-4k-drone-photo-dataset,
  title = { Visualization Tools for NTUT 4K Drone Photo Dataset },
  type = { Computer Vision Tools },
  author = { Dataset Ninja },
  howpublished = { \url{ https://datasetninja.com/ntut-4k-drone-photo } },
  url = { https://datasetninja.com/ntut-4k-drone-photo },
  journal = { Dataset Ninja },
  publisher = { Dataset Ninja },
  year = { 2024 },
  month = { jul },
  note = { visited on 2024-07-27 },
}

Download #

Please visit dataset homepage to download the data.

. . .

Disclaimer #

Our gal from the legal dep told us we need to post this:

Dataset Ninja provides visualizations and statistics for some datasets that can be found online and can be downloaded by general audience. Dataset Ninja is not a dataset hosting platform and can only be used for informational purposes. The platform does not claim any rights for the original content, including images, videos, annotations and descriptions. Joint publishing is prohibited.

You take full responsibility when you use datasets presented at Dataset Ninja, as well as other information, including visualizations and statistics we provide. You are in charge of compliance with any dataset license and all other permissions. You are required to navigate datasets homepage and make sure that you can use it. In case of any questions, get in touch with us at hello@datasetninja.com.