The MS COCO (Microsoft Common Objects in Context) dataset is a large-scale object detection, segmentation, key-point detection, and captioning dataset.

COCO Dataset Limited (Person Only)

## Overview

COCO Dataset Limited (Person Only) is a dataset for object detection tasks - it contains People annotations for 5,438 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

This is the full 2017 COCO object detection dataset (train and valid), which is a subset of the most recent 2020 COCO object detection dataset.

COCO is a large-scale object detection, segmentation, and captioning dataset of many object types easily recognizable by a 4-year-old. The data is initially collected and published by Microsoft. The original source of the data is here and the paper introducing the COCO dataset is here.

Please note: this archive requires support for dangling symlinks, which excludes the Windows operating system.

To use this dataset, you will need to download the MS COCO 2017 detection images and expand them to a folder called coco17 in the train_val_combined directory. The download can be found here: https://cocodataset.org/#download You will also need to download the AI2D image description dataset and expand them to a folder called ai2d in the train_val_combined directory. The download can be found here: https://prior.allenai.org/projects/diagram-understanding

License Notes for Train and Val: Since the images in this dataset come from different sources, they are bound by different licenses.

Images for bar charts, x-y plots, maps, pie charts, tables, and technical drawings were downloaded directly from wikimedia commons. License and authorship information is stored independently for each image in these categories in the wikimedia_commons_licenses.csv file. Each row (note: some rows are multi-line) is formatted so:

Images in the slides category were taken from presentations which were downloaded from Wikimedia Commons. The names of the presentations on Wikimedia Commons omits the trailing underscore, number, and file extension, and ends with .pdf instead. The source materials' licenses are shown in source_slices_licenses.csv.

Wikimedia commons photos' information page can be found at "https://commons.wikimedia.org/wiki/File:

License Notes for Testing: The testing images have been uploaded to SlideWiki by SlideWiki users. The image authorship and copyright information is available in authors.csv.

Further information can be found for each image using the SlideWiki file service. Documentation is available at https://fileservice.slidewiki.org/documentation#/ and in particular: metadata is available at "https://fileservice.slidewiki.org/metadata/

This is the SlideImages dataset, which has been assembled for the SlideImages paper. If you find the dataset useful, please cite our paper: https://doi.org/10.1007/978-3-030-45442-5_36

Microsoft Common Objects in Context (COCO) Dataset

The Common Objects in Context (COCO) dataset is a widely recognized collection designed to spur object detection, segmentation, and captioning research. Created by Microsoft, COCO provides annotations, including object categories, keypoints, and more. The model it a valuable asset for machine learning practitioners and researchers. Today, many model architectures are benchmarked against COCO, which has enabled a standard system by which architectures can be compared.

While COCO is often touted to comprise over 300k images, it's pivotal to understand that this number includes diverse formats like keypoints, among others. Specifically, the labeled dataset for object detection stands at 123,272 images.

The full object detection labeled dataset is made available here, ensuring researchers have access to the most comprehensive data for their experiments. With that said, COCO has not released their test set annotations, meaning the test data doesn't come with labels. Thus, this data is not included in the dataset.

The Roboflow team has worked extensively with COCO. Here are a few links that may be helpful as you get started working with this dataset:

• An introduction to the COCO dataset
• Weird images in COCO, and what that tells us about the utility and limits of COCO

Mixed Image-Caption Dataset (COCO2014 + CC12M)

This dataset contains a collection of 32,000 image-caption pairs sourced from:

• COCO 2014: https://cocodataset.org/#home
• CC12M (Conceptual Captions 12M): https://github.com/google-research-datasets/conceptual-12m

Each entry is included in the JSON file train_mix_32000.json, with the following fields: - "filename": Image filename (relative to dataset structure) - "caption": Image description - "data": Source dataset ("coco" or "cc12m")

📦 Included

• train_mix_32000.json: Metadata file with image paths and captions.
• images/: Folder structure containing all 32,000 actual image files referenced in the JSON.

💡 Image paths in the JSON have been adjusted to reflect the folder structure inside this Kaggle dataset.

📄 License

This dataset includes images from:

• COCO 2014
  Licensed under Creative Commons Attribution 4.0.

• CC12M
  Provided by Google LLC under a permissive license:

  The dataset may be freely used for any purpose, although acknowledgment of Google LLC as the data source would be appreciated.
  The dataset is provided "AS IS" without any warranty, express or implied.
  View License

🧠 Use Cases

• Vision-language pretraining
• Knowledge-enhanced captioning
• Image-text retrieval tasks
• Multi-task learning in vision-language models

🙏 Acknowledgements

• COCO Dataset
• Google Conceptual Captions (CC12M)

Credit belongs to https://cocodataset.org

coco2017

Image-text pairs from MS COCO2017.

  Data origin

Data originates from cocodataset.org While coco-karpathy uses a dense format (with several sentences and sendids per row), coco-karpathy-long uses a long format with one sentence (aka caption) and sendid per row. coco-karpathy-long uses the first five sentences and therefore is five times as long as coco-karpathy. phiyodr/coco2017: One row corresponds one image with several sentences. phiyodr/coco2017-long: One row… See the full description on the dataset page: https://huggingface.co/datasets/phiyodr/coco2017.

Dataset: COCO-Formatted Object Detection Dataset

Overview

This dataset is designed for object detection tasks and follows the COCO format. It contains 300 images and corresponding annotation files in JSON format. The dataset is split into training, validation, and test sets, ensuring a balanced distribution for model evaluation.

Dataset Structure

The dataset is organized into three main folders:

train/ (70% - 210 images)

valid/ (15% - 45 images)

test/ (15% - 45 images)

Each folder contains:

Images in JPEG/PNG format.

A corresponding _annotations.coco.json file that includes bounding box annotations.

Preprocessing & Augmentations

The dataset has undergone several preprocessing and augmentation steps to enhance model generalization:

Image Preprocessing:

Auto-orientation applied

Resized to 640x640 pixels (stretched)

Augmentation Techniques:

Flip: Horizontal flipping

Crop: 0% minimum zoom, 5% maximum zoom

Rotation: Between -5° and +5°

Saturation: Adjusted between -4% and +4%

Brightness: Adjusted between -10% and +10%

Blur: Up to 0px

Noise: Up to 0.1% of pixels

Bounding Box Augmentations:

Flipping, cropping, rotation, brightness adjustments, blur, and noise applied accordingly to maintain annotation consistency.

Annotation Format

The dataset follows the COCO (Common Objects in Context) format, which includes:

images section: Contains image metadata such as filename, width, and height.

annotations section: Includes bounding boxes, category IDs, and segmentation masks (if applicable).

categories section: Defines class labels.

COCO Dataset From YOLO

## Overview

COCO Dataset From YOLO is a dataset for object detection tasks - it contains Objects annotations for 9,330 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

This dataset contains all COCO 2017 images and annotations split in training (118287 images) and validation (5000 images).

Real-world dataset of ~400 images of cuboid-shaped parcels with full 2D and 3D annotations in the COCO format.

Relevant computer vision tasks:

• bounding box detection
• instance segmentation
• keypoint estimation
• 3D bounding box estimation
• 3D voxel reconstruction (.binvox files)
• 3D reconstruction (.obj files)

For details, see our paper and project page.

If you use this resource for scientific research, please consider citing

@inproceedings{naumannScrapeCutPasteLearn2022,
  title    = {Scrape, Cut, Paste and Learn: Automated Dataset Generation Applied to Parcel Logistics},
  author    = {Naumann, Alexander and Hertlein, Felix and Zhou, Benchun and Dörr, Laura and Furmans, Kai},
  booktitle  = {{{IEEE Conference}} on {{Machine Learning}} and Applications ({{ICMLA}})},
  date     = 2022
}

Mini COCO Dataset

## Overview

Mini COCO Dataset is a dataset for instance segmentation tasks - it contains People annotations for 493 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

Original Coco

## Overview

Original Coco is a dataset for object detection tasks - it contains Teeth NJqp annotations for 676 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

Synthetic dataset of over 13,000 images of damaged and intact parcels with full 2D and 3D annotations in the COCO format. For details see our paper and for visual samples our project page.

Relevant computer vision tasks:

• bounding box detection
• classification
• instance segmentation
• keypoint estimation
• 3D bounding box estimation
• 3D voxel reconstruction
• 3D reconstruction

The dataset is for academic research use only, since it uses resources with restrictive licenses.
For a detailed description of how the resources are used, we refer to our paper and project page.

Licenses of the resources in detail:

• Google Scanned Objects: CC BY 4.0 (for details on which files are used, see the respective meta folder)
• Cardboard Dataset: CC BY 4.0
• Shipping Label Dataset: CC BY-NC 4.0
• Other Labels: See file misc/source_urls.json
• LDR Dataset: License for Non-Commercial Use
• Large Logo Dataset (LLD): Please notice that this dataset is made available for academic research purposes only. All the images are collected from the Internet, and the copyright belongs to the original owners. If any of the images belongs to you and you would like it removed, please kindly inform us, we will remove it from our dataset immediately.

You can use our textureless models (i.e. the obj files) of damaged parcels under CC BY 4.0 (note that this does not apply to the textures).

If you use this resource for scientific research, please consider citing

@inproceedings{naumannParcel3DShapeReconstruction2023,
  author  = {Naumann, Alexander and Hertlein, Felix and D\"orr, Laura and Furmans, Kai},
  title   = {Parcel3D: Shape Reconstruction From Single RGB Images for Applications in Transportation Logistics},
  booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops},
  month   = {June},
  year   = {2023},
  pages   = {4402-4412}
}

Dataset Overview

This page contains a modified Cocos dataset along with details about the dataset used.

File Descriptions

imgs.zip - Train: 🚂 This folder contains the training set, which can be split into train/validation data for model training. - Test: 🧪 Your trained models should be used to produce predictions on the test set.

labels.zip - categories.csv: 📝 This file lists all the object classes in the dataset, ordered according to the column ordering in the train labels file. - train_labels.csv: 📊 This file contains data regarding which image contains which categories.

Dataset The Man OverBoard Drone (MOBDrone) dataset is a large-scale collection of aerial footage images. It contains 126,170 frames extracted from 66 video clips gathered from one UAV flying at an altitude of 10 to 60 meters above the mean sea level. Images are manually annotated with more than 180K bounding boxes localizing objects belonging to 5 categories --- person, boat, lifebuoy, surfboard, wood. More than 113K of these bounding boxes belong to the person category and localize people in the water simulating the need to be rescued. In this repository, we provide: 66 Full HD video clips (total size: 5.5 GB) 126,170 images extracted from the videos at a rate of 30 FPS (total size: 243 GB) 3 annotation files for the extracted images that follow the MS COCO data format (for more info see https://cocodataset.org/#format-data): annotations_5_custom_classes.json: this file contains annotations concerning all five categories; please note that class ids do not correspond with the ones provided by the MS COCO standard since we account for two new classes not previously considered in the MS COCO dataset --- lifebuoy and wood annotations_3_coco_classes.json: this file contains annotations concerning the three classes also accounted by the MS COCO dataset --- person, boat, surfboard. Class ids correspond with the ones provided by the MS COCO standard. annotations_person_coco_classes.json: this file contains annotations concerning only the 'person' class. Class id corresponds to the one provided by the MS COCO standard. The MOBDrone dataset is intended as a test data benchmark. However, for researchers interested in using our data also for training purposes, we provide training and test splits: Test set: All the images whose filename starts with "DJI_0804" (total: 37,604 images) Training set: All the images whose filename starts with "DJI_0915" (total: 88,568 images) More details about data generation and the evaluation protocol can be found at our MOBDrone paper: https://arxiv.org/abs/2203.07973 The code to reproduce our results is available at this GitHub Repository: https://github.com/ciampluca/MOBDrone_eval See also http://aimh.isti.cnr.it/dataset/MOBDrone Citing the MOBDrone The MOBDrone is released under a Creative Commons Attribution license, so please cite the MOBDrone if it is used in your work in any form. Published academic papers should use the academic paper citation for our MOBDrone paper, where we evaluated several pre-trained state-of-the-art object detectors focusing on the detection of the overboard people @inproceedings{MOBDrone2021, title={MOBDrone: a Drone Video Dataset for Man OverBoard Rescue}, author={Donato Cafarelli and Luca Ciampi and Lucia Vadicamo and Claudio Gennaro and Andrea Berton and Marco Paterni and Chiara Benvenuti and Mirko Passera and Fabrizio Falchi}, booktitle={ICIAP2021: 21th International Conference on Image Analysis and Processing}, year={2021} } and this Zenodo Dataset @dataset{donato_cafarelli_2022_5996890, author={Donato Cafarelli and Luca Ciampi and Lucia Vadicamo and Claudio Gennaro and Andrea Berton and Marco Paterni and Chiara Benvenuti and Mirko Passera and Fabrizio Falchi}, title = {{MOBDrone: a large-scale drone-view dataset for man overboard detection}}, month = feb, year = 2022, publisher = {Zenodo}, version = {1.0.0}, doi = {10.5281/zenodo.5996890}, url = {https://doi.org/10.5281/zenodo.5996890} } Personal works, such as machine learning projects/blog posts, should provide a URL to the MOBDrone Zenodo page (https://doi.org/10.5281/zenodo.5996890), though a reference to our MOBDrone paper would also be appreciated. Contact Information If you would like further information about the MOBDrone or if you experience any issues downloading files, please contact us at mobdrone[at]isti.cnr.it Acknowledgements This work was partially supported by NAUSICAA - "NAUtical Safety by means of Integrated Computer-Assistance Appliances 4.0" project funded by the Tuscany region (CUP D44E20003410009). The data collection was carried out with the collaboration of the Fly&Sense Service of the CNR of Pisa - for the flight operations of remotely piloted aerial systems - and of the Institute of Clinical Physiology (IFC) of the CNR - for the water immersion operations.

The Dataset

A collection of images of parking lots for vehicle detection, segmentation, and counting.
Each image is manually labeled with pixel-wise masks and bounding boxes localizing vehicle instances.
The dataset includes about 250 images depicting several parking areas describing most of the problematic situations that we can find in a real scenario: seven different cameras capture the images under various weather conditions and viewing angles. Another challenging aspect is the presence of partial occlusion patterns in many scenes such as obstacles (trees, lampposts, other cars) and shadowed cars.
The main peculiarity is that images are taken during the day and the night, showing utterly different lighting conditions.

We suggest a three-way split (train-validation-test). The train split contains images taken during the daytime while validation and test splits include images gathered at night.
In line with these splits we provide some annotation files:

• train_coco_annotations.json and val_coco_annotations.json --> JSON files that follow the golden standard MS COCO data format (for more info see https://cocodataset.org/#format-data) for the training and the validation splits, respectively. All the vehicles are labeled with the COCO category 'car'. They are suitable for vehicle detection and instance segmentation.

• train_dot_annotations.csv and val_dot_annotations.csv --> CSV files that contain xy coordinates of the centroids of the vehicles for the training and the validation splits, respectively. Dot annotation is commonly used for the visual counting task.

• ground_truth_test_counting.csv --> CSV file that contains the number of vehicles present in each image. It is only suitable for testing vehicle counting solutions.

Citing our work

If you found this dataset useful, please cite the following paper

@inproceedings{Ciampi_visapp_2021,
  doi = {10.5220/0010303401850195},
  url = {https://doi.org/10.5220%2F0010303401850195},
  year = 2021,
  publisher = {{SCITEPRESS} - Science and Technology Publications},
  author = {Luca Ciampi and Carlos Santiago and Joao Costeira and Claudio Gennaro and Giuseppe Amato},
  title = {Domain Adaptation for Traffic Density Estimation},
  booktitle = {Proceedings of the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications}
}

and this Zenodo Dataset

@dataset{ciampi_ndispark_6560823,
  author = {Luca Ciampi and Carlos Santiago and Joao Costeira and Claudio Gennaro and Giuseppe Amato},
  title = {{Night and Day Instance Segmented Park (NDISPark) Dataset: a Collection of Images taken by Day and by Night for Vehicle Detection, Segmentation and Counting in Parking Areas}},
  month = may,
  year = 2022,
  publisher = {Zenodo},
  version = {1.0.0},
  doi = {10.5281/zenodo.6560823},
  url = {https://doi.org/10.5281/zenodo.6560823}
}

Contact Information

If you would like further information about the dataset or if you experience any issues downloading files, please contact us at luca.ciampi@isti.cnr.it

MS COCO is a large-scale object detection, segmentation, and captioning dataset. COCO has several features: Object segmentation, Recognition in context, Superpixel stuff segmentation, 330K images (>200K labeled), 1.5 million object instances, 80 object categories, 91 stuff categories, 5 captions per image, 250,000 people with keypoints.

Dataset

This dataset was created by Hai Vo

Contents