glenn-jocher
commented
2 years ago @bryanbo-cao 👋 Hello! Thanks for asking about YOLOv5 🚀 dataset formatting. You could just use one data.yaml and bash script to rename your directories between each of the 16 trainings.
For examples of using image directories instead of txt lists of images see other datasets like VOC.yaml: https://github.com/ultralytics/yolov5/blob/d6051382f1551455b88ca086b99275cfc8286131/data/VOC.yaml#L1-L21
To train correctly your data must be in YOLOv5 format. Please see our Train Custom Data tutorial for full documentation on dataset setup and all steps required to start training your first model. A few excerpts from the tutorial:
1.1 Create dataset.yaml
COCO128 is an example small tutorial dataset composed of the first 128 images in COCO train2017. These same 128 images are used for both training and validation to verify our training pipeline is capable of overfitting. data/coco128.yaml, shown below, is the dataset config file that defines 1) the dataset root directory path and relative paths to train / val / test image directories (or *.txt files with image paths), 2) the number of classes nc and 3) a list of class names:
# Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..]
path: ../datasets/coco128 # dataset root dir
train: images/train2017 # train images (relative to 'path') 128 images
val: images/train2017 # val images (relative to 'path') 128 images
test: # test images (optional)
# Classes
nc: 80 # number of classes
names: [ 'person', 'bicycle', 'car', 'motorcycle', 'airplane', 'bus', 'train', 'truck', 'boat', 'traffic light',
'fire hydrant', 'stop sign', 'parking meter', 'bench', 'bird', 'cat', 'dog', 'horse', 'sheep', 'cow',
'elephant', 'bear', 'zebra', 'giraffe', 'backpack', 'umbrella', 'handbag', 'tie', 'suitcase', 'frisbee',
'skis', 'snowboard', 'sports ball', 'kite', 'baseball bat', 'baseball glove', 'skateboard', 'surfboard',
'tennis racket', 'bottle', 'wine glass', 'cup', 'fork', 'knife', 'spoon', 'bowl', 'banana', 'apple',
'sandwich', 'orange', 'broccoli', 'carrot', 'hot dog', 'pizza', 'donut', 'cake', 'chair', 'couch',
'potted plant', 'bed', 'dining table', 'toilet', 'tv', 'laptop', 'mouse', 'remote', 'keyboard', 'cell phone',
'microwave', 'oven', 'toaster', 'sink', 'refrigerator', 'book', 'clock', 'vase', 'scissors', 'teddy bear',
'hair drier', 'toothbrush' ] # class names1.2 Create Labels
After using a tool like Roboflow Annotate to label your images, export your labels to YOLO format, with one *.txt file per image (if no objects in image, no *.txt file is required). The *.txt file specifications are:
- One row per object
- Each row is
class x_center y_center width heightformat. - Box coordinates must be in normalized xywh format (from 0 - 1). If your boxes are in pixels, divide
x_centerandwidthby image width, andy_centerandheightby image height. - Class numbers are zero-indexed (start from 0).
The label file corresponding to the above image contains 2 persons (class 0) and a tie (class 27):
1.3 Organize Directories
Organize your train and val images and labels according to the example below. YOLOv5 assumes /coco128 is inside a /datasets directory next to the /yolov5 directory. YOLOv5 locates labels automatically for each image by replacing the last instance of /images/ in each image path with /labels/. For example:
../datasets/coco128/images/im0.jpg # image
../datasets/coco128/labels/im0.txt # label
Good luck 🍀 and let us know if you have any other questions!
bryanbocao
github-actions[bot]
GusevMihail
chobits
Akshaykushawaha
Search before asking
Question
Hello! I like the way this repo organize! I was trying to do some sort of "grid search" for investigating performance of Yolo. Specifically, I have a
basecoco dataset, the one exactly downloaded by the script incoco.yamland would like to have variations on two levels: (1) in the image data level, I do some image processing and have different sets of image, sayimages_v2, images_v3, images_v4whileimagesis the base one; (2) in the label level for bbox, I also have different variations such as changing label names, number of classes or category ids saved in various sets of label folders:labels_v2, labels_v3, labels_v4.Below is a brief structure of files in
dataset/coco:By "Grid Search" I mean I will have one result for each pair of
images*andlabels*, resulting in4(images) x 4(labels) =16sets of experiments in total.Q1: Is there any way to do that efficiently?
A straight forward way is to have
16datasets ofcocolikecoco_1,coco_2,coco_3while each corresponds to one pair ofimages*andlabels*. However, it requires16 x 20.1GB=321.6GBspace which is too much for me.When sweeping
images*, it seems that I can just change the image paths intrain2017.txtandval2017.txt, but the default label path islabelsand I don't see I can specify the path in https://github.com/ultralytics/yolov5/blob/master/data/coco.yaml. Q2: Is there any way to do that?Appreciate your help!
Additional
No response