Skip to content

Dataset · rafaelpadilla

COCO 2017 Object Detection Mirror

rafaelpadilla/coco2017

A complete set of COCO 2017 images and annotations for training and evaluating computer vision models.

What's inside

This collection contains all images and annotations from the classic COCO 2017 benchmark. The data is split into a training set of 118,287 images and a validation set of 5,000 images, complete with English captions. Judging by the repository name, this is a user re-upload of the well-known corpus.

What you can do with it

  • Training and fine-tuning object detection models (YOLO, DETR)
  • Validating and comparing the accuracy of computer vision algorithms on a standard benchmark
Tasks
Object detection
Languages
English
Size
100K–1M rows
Downloads this month
13.6K
Likes
32

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("rafaelpadilla/coco2017")