CODEWITHSUNDEEP
yolov8posecvat
user28718285
user28718285

Reputation: 11

Issue importing yoloV8 pose annotations, images load annotations do not

I am attempting to import images and annotations in YoloV8 pose format. The problem is when I try to upload my annotations the images load but there are no labeled points loaded.

I have my files organized in the same way as in this example:

https://docs.cvat.ai/docs/manual/advanced/formats/format-yolov8/#:~:text=YOLOv8%20is%20a%20format%20family,Pose

I have 355 images and in each image is roughly 50 sharks, each shark has 6 key points

my data.yaml looks like this:

data.yaml:

path: ./ # dataset root dir train: images/train/ # train images (relative to 'path')

Keypoint settings

kpt_shape: [6, 2]

Classes

names: 0: shark

my train.txt contains a row for each image that looks like: images/train/frame1.png

The steps I am following right now are to

  1. create a new task
  2. add a single label called "shark" as I only have one class
  3. I drag and drop my zipped folder that contains the images/train and labels/train subfolders, as well as the data.yaml and the train.txt
  4. I click submit and open

The result is that the images will appear, but no labeled points or bounding boxes show up in the job.

Am I missing a step or doing something wrong? Happy to provide more info if needed. Thanks in advance -Darien

Upvotes: 1

Views: 25

Answers (0)

Related Questions