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

MS-COCO2017

  Use the dataset

from random import randint from datasets import load_dataset from PIL import Image, ImageDraw, ImageFont

ds = load_dataset("ariG23498/coco2017", streaming=True, split="validation")

sample = next(iter(ds))

def draw_bboxes_on_image( image: Image.Image, objects: dict, category_names: dict = None, box_color: str = "red", text_color: str = "white" ) -> Image.Image: image_copy = image.copy() draw =… See the full description on the dataset page: https://huggingface.co/datasets/ariG23498/coco2017.

Purpose: experiments with YOLO models in monochrome.

The original COCO2017 dataset has been processed: - added YOLO annotations for 80 classes; - all images are converted to monochrome (greyscale) with an equalized histogram.

The number of images: - training: 118,287; - validation: 5,000.

Links to the original COCO 2017 dataset https://cocodataset.org by Microsoft: url_images = 'http://images.cocodataset.org/zips/' url_annotations = 'http://images.cocodataset.org/annotations/annotations_trainval2017.zip'

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.

MicroCOCO2017 is a curated subset of the COCO 2017 dataset, designed for lightweight experimentation with object detection and segmentation models. It includes: - 25,000 images from the train2017 split - 5,000 images from the val2017 split Full COCO-style annotations (bounding boxes, categories, segmentation masks) This dataset is ideal for faster training and prototyping while maintaining the diversity of the original COCO dataset. 📁 Forked from: github.com/giddyyupp/coco-minitrain

I wanted to train a custom YOLO object detection model, but the MS-COCO dataset was not in a good format. So I parsed the instances json files in the MS-COCO annotations and processed the dataset to be a YOLO friendly format.

I downloaded the dataset from COCO webste. You can download any split you need from the COCO dataset website

Directory info: 1. test: Only contains the test images 2. train: Has two sub folders, images - contains the training images, labels - contains the training labels in a .txt file for each train image 3. val: Has two sub folders, images - contains the validation images, labels - contains the validation labels in a .txt file for each validation image

I do not own the dataset in any way. I merely parsed the dataset to a be in a ready to train YOLO format. Download the original dataset from the COCO webste

COCO 2017 Dataset for Image Colorization

  Overview

This dataset is derived from the COCO (Common Objects in Context) 2017 dataset, which is a large-scale object detection, segmentation, and captioning dataset. The COCO 2017 dataset has been adapted here specifically for the task of image colorization.

  Format

DatasetDict({ train: Dataset({ features: ['license', 'file_name', 'coco_url', 'height', 'width', 'date_captured', 'flickr_url', 'image_id'… See the full description on the dataset page: https://huggingface.co/datasets/nickpai/coco2017-colorization.

Coco2017

## Overview

Coco2017 is a dataset for object detection tasks - it contains Images annotations for 7,557 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).

Ms Coco 2017

## Overview

Ms Coco 2017 is a dataset for object detection tasks - it contains Ms Coco 2017 annotations for 4,611 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 [Public Domain license](https://creativecommons.org/licenses/Public Domain).

Dataset

This dataset was created by Yurii Pryimak

Contents

Coco2017 Person Ver

## Overview

Coco2017 Person Ver is a dataset for object detection tasks - it contains Person annotations for 900 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).

Object detection is one of the most foundational computer vision task and is essential for many real-world applications. The object detection pipeline has been developed rapidly, especially in the era of deep learning.

📌 What's Included:

• Training Set: 118K images with annotations for detection, segmentation, and keypoints.
• Validation Set: 5K images with full annotations for validation.
• Testing Set: Images are divided into two splits—dev and challenge—replacing the four splits (dev, standard, reserve, challenge) used in previous years.
• Stuff Annotations: Available for 40K images in the training set and 5K validation images, enabling semantic segmentation research.
• Unlabeled Data: A set of 120K images with no annotations, mirroring the class distribution of the labeled data. This is ideal for exploring semi-supervised learning techniques.

🔍 Key Changes in COCO 2017:

• The train/val split was updated based on community feedback, now featuring 118K/5K images instead of the previous 83K/41K split.
• While the annotations for detection and keypoints are consistent with previous years, additional stuff annotations were introduced in 2017.
• Unlabeled data is now available for semi-supervised learning tasks, opening new avenues for experimentation.

📂 Dataset Structure:

• train2017: Images and annotations
• val2017: Images and annotations
• test2017: Images (no annotations provided)
• unlabeled2017: Unlabeled images

This dataset can be used for a variety of computer vision tasks, including object detection, instance segmentation, keypoint detection, semantic segmentation, and image captioning. Whether you're working on supervised or semi-supervised learning, this resource is designed to meet your needs.

📄 License and Attribution

This dataset is a downsampled version of the COCO 2017 dataset, tailored for segmentation tasks. It has the following fields:

image: 256x256 image segmentation: 256x256 image. Each pixel encodes the class of that pixel. See class_names_dict.json for a legend. captions: a list of captions for the image, each by a different labeler.

Use the dataset as follows: import requests from datasets import load_dataset

ds =… See the full description on the dataset page: https://huggingface.co/datasets/peteole/coco2017-segmentation-10k-256x256.

COCO2017 No Crowd Only People

## Overview

COCO2017 No Crowd Only People is a dataset for instance segmentation tasks - it contains Person annotations for 5,393 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).

COCO2017 0 ~ 1000 Person Ver

## Overview

COCO2017 0 ~ 1000 Person Ver is a dataset for object detection tasks - it contains Person annotations for 900 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 COCO-2017 dataset's training split for object detection and segmentation saved in the Lance file format for blazing fast and memory-efficient I/O.

This dataset only includes data necessary for object detection and segmentation.

For detailed information on how the dataset was created, refer to my article on Curating Custom Datasets for efficient LLM training using Lance.

Instructions for using this dataset

This dataset is not supposed to be used on Kaggle Kernels since Lance requires the input directory of the dataset to have write access but Kaggle Kernel's input directory doesn't have it and the dataset size prohibits one from moving it to /kaggle/working. Hence, to use this dataset, you must download it by using the Kaggle API or through this page, and then move the unzipped files to a folder called coco2017_train.lance. Below are detailed snippets on how to download and use this dataset.

First, download and unzip the dataset from your terminal (make sure you have your kaggle API key at ~/.kaggle/:

$ pip install -q kaggle pyarrow pylance
$ kaggle datasets download -d heyytanay/coco2017-train-lance
$ mkdir coco2017_train.lance/
$ unzip -qq coco2017-train-lance.zip -d coco2017_train.lance/
$ rm coco2017-train-lance.zip

Once this is done, you will find your dataset in the coco2017_train.lance/ folder. To load and get the gist of the data, run the below snippet.

import lance
dataset = lance.dataset('coco2017_train.lance/')
print(dataset.count_rows())

This will give you the total number of rows in the dataset.

The MS-COCO 2017 test-dev17 dataset is used to evaluate the performances of various RefineDet networks.

Taoyang1/COCO2017 dataset hosted on Hugging Face and contributed by the HF Datasets community

COCO Object Detection Dataset | 2017

Downloaded from here and it includes Train images for now.