Shulou(Shulou.com)05/31 Report--

This article mainly introduces "how to use pytorch to load and read COCO datasets". In daily operation, I believe many people have doubts about how to use pytorch to load and read COCO datasets. I have consulted all kinds of materials and sorted out simple and easy operation methods. I hope to help you answer the doubts about "how to use pytorch to load and read COCO datasets"! Next, please follow the small series to learn together!

environment configuration

Watch the pytorch tutorial

Basics: Yuanzu, Dictionary, Array #Yuanzu a = (1, 2)#Dictionary b = {'username': 'peipeiwang',' code':'111'}#Array c = [1, 2, 3]print(a[0])print(c[0])print(b["username"])

Output:

Use PyTorch to read COCO dataset import torchvisionfrom PIL import ImageDraw#import coco 2017 validation set and corresponding annotationscoco_dataset = torchvision.datasets.CocoDetection(root="COCO_dataset_val_2017/val2017", annFile="COCO_dataset_val_2017/annotations_trainval2017/annotations/instances_val2017.json")#Read image and annotation separately image, info = coco_dataset[0]# ImageDraw drawing tool image_handler = ImageDraw.ImageDraw(image)for annotation in info: # bbox is the position coordinate of the detection box x_min, y_min, width, height = annotation['bbox'] # ((), ()) are the coordinate pair of the upper left corner and the coordinate pair of the upper right corner respectively, image_handler.rectangle refers to the drawing box in the picture image_handler.rectangle(((x_min, y_min), (x_min + width, y_min + height)))image.show()

Results:

Using PyTorch to read your own datasets

Create your own dataset labels using the cvat tool, export to coco format and read

Results:

At this point, the study of "how to load and read COCO datasets using pytorch" is over, hoping to solve everyone's doubts. Theory and practice can better match to help everyone learn, go and try it! If you want to continue learning more relevant knowledge, please continue to pay attention to the website, Xiaobian will continue to strive to bring more practical articles for everyone!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.