Why divide by 1000 when calculating inference time per image?

[Jungjihyuk]

Search before asking

• I have searched the YOLOv5 issues and discussions and found no similar questions.

Question

Hello,
I have a question about inference time per image.

detect.py code in YOLOv5 v7.0

line 207~209

# Print results t = tuple(x.t / seen * 1E3 for x in dt) # speeds per image LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms NMS per image at shape {(1, 3, *imgsz)}' % t)

Question
If x.t is in seconds and the logger is in ms then shouldn't it be multiplied by 1000?
For example, if x.t is 0.2s and convert to ms, it is 20ms.
So we should do 0.2*1000.
please let me know if i missed out anything.
Thank you.

Additional

No response

[github-actions]

👋 Hello @Jungjihyuk, thank you for your interest in YOLOv5 🚀! Please visit our ⭐️ Tutorials to get started, where you can find quickstart guides for simple tasks like Custom Data Training all the way to advanced concepts like Hyperparameter Evolution.

If this is a 🐛 Bug Report, please provide a minimum reproducible example to help us debug it.

If this is a custom training ❓ Question, please provide as much information as possible, including dataset image examples and training logs, and verify you are following our Tips for Best Training Results.

Requirements

Python>=3.7.0 with all requirements.txt installed including PyTorch>=1.7. To get started:

git clone https://github.com/ultralytics/yolov5  # clone
cd yolov5
pip install -r requirements.txt  # install

Environments

YOLOv5 may be run in any of the following up-to-date verified environments (with all dependencies including CUDA/CUDNN, Python and PyTorch preinstalled):

• Notebooks with free GPU:
• Google Cloud Deep Learning VM. See GCP Quickstart Guide
• Amazon Deep Learning AMI. See AWS Quickstart Guide
• Docker Image. See Docker Quickstart Guide

Status

If this badge is green, all YOLOv5 GitHub Actions Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 training, validation, inference, export and benchmarks on macOS, Windows, and Ubuntu every 24 hours and on every commit.

Introducing YOLOv8 🚀

We're excited to announce the launch of our latest state-of-the-art (SOTA) object detection model for 2023 - YOLOv8 🚀!

Designed to be fast, accurate, and easy to use, YOLOv8 is an ideal choice for a wide range of object detection, image segmentation and image classification tasks. With YOLOv8, you'll be able to quickly and accurately detect objects in real-time, streamline your workflows, and achieve new levels of accuracy in your projects.

Check out our YOLOv8 Docs for details and get started with:

pip install ultralytics

[glenn-jocher]

Hello @Jungjihyuk,

Thank you for reaching out to us with your question.

Regarding your question, detect.py code in YOLOv5 v7.0 multiplies the speeds per image values by 1E3 for conversion from seconds to milliseconds. Specifically, tuple(x.t / seen * 1E3 for x in dt) computes the average speeds per image as tuple (pre-process time, inference time, NMS time) in milliseconds.

Therefore, you don't need to multiply the calculation of milliseconds by 1000 when logging the inference time per image.

Please let us know if you have further questions. We're always here to assist you.

Best,
Ultralytics Team

[Jungjihyuk]

Thanks for your reply.

After reading the answer carefully, I realized that it was my mistake.
I misunderstood x.t / seen * 1E3 code as x.t /(seen * 1E3)
Now i totally understand.

Thank you glenn.

[glenn-jocher]

Dear @Jungjihyuk,

Thank you for getting back to us with your feedback, and we're delighted to know that the issue is resolved.

Please don't hesitate to ask if you have any further questions or need further assistance.

Best,
Glenn Jocher - Ultralytics Team