Hi there, I’m a highschool student trying to create a tracking software with YOLO AI for daphnia movement. Currently I’m testing my program with a datapool of 60 pictures from a 1 min video. (1 picture per 1 second) However, I can’t seem to get the YOLO program to recognize the labels. Can anyone suggest ideas on what to do? I’ve already tried to troubleshoot by checking the labels to see if they have the correct information, checking if the names of the images match the labels and checking if my YAML files are correct.
This is my files setup:
C:/Daphnia_Project/Dataset_Images/
├── daphnia.yaml
├── Images/
│ ├── train/
│ │ ├── xxx.jpg
│ └── val/
│ ├── xxx.jpg
├── Labels/
│ ├── train/
│ │ ├── xxx.txt
│ └── val/
│ ├── xxx.txt
“train” has 48 images and texts while “val” has 12.
This is my YAML file:
train: C:/Daphnia_Project/Dataset_Images/Images/train
val: C:/Daphnia_Project/Dataset_Images/Images/val
names:
0: daphnia
This is my python code to fun YOLO training:
from ultralytics import YOLO
# Loading small pretrained model
model = YOLO(“yolov8n.pt”)
# Training the model
model.train(
data=“C:/Daphnia_Project/Dataset_Images/daphnia.yaml”, #path to YAML
epochs=100,
imgsz=1536,
batch=8,
#Small object tuning
degrees=0.0,
scale=0.5,
mosaic=1.0,
close_mosaic=10,
hsv_h=0.015,
hsv_s=0.7,
hsv_v=0.4,
# Save settings
project=“C:/Daphnia_Project/runs”,
name=“daphnia_model”,
pretrained=True
)
This is the error that constantly appears:
WARNING no labels found in detect set, cannot compute metrics without labels