AP calculation in metrics.py

[yusiyoh]

❔Question

Why do not you use mrec = np.concatenate(([0.], recall, [1.])) instead of

yolov5/utils/metrics.py

Line 94 in 2da6444

mrec = np.concatenate(([0.], recall, [recall[-1] + 0.01]))

? It changes the mAP results. I could not get the idea behind it.

Additional context

[github-actions]

👋 Hello @yusiyoh, 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 screenshots and minimum viable code to reproduce your issue, otherwise we can not help you.

If this is a custom training ❓ Question, please provide as much information as possible, including dataset images, training logs, screenshots, and a public link to online W&B logging if available.

For business inquiries or professional support requests please visit https://ultralytics.com or email Glenn Jocher at glenn.jocher@ultralytics.com.

Requirements

Python>=3.6.0 with all requirements.txt installed including PyTorch>=1.7. To get started:

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

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):

• Google Colab and Kaggle 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 (train.py), validation (val.py), inference (detect.py) and export (export.py) on MacOS, Windows, and Ubuntu every 24 hours and on every commit.

[glenn-jocher]

@yusiyoh hi, thank you for your feature suggestion on how to improve YOLOv5 🚀!

The fastest and easiest way to incorporate your ideas into the official codebase is to submit a Pull Request (PR) implementing your idea, and if applicable providing before and after profiling/inference/training results to help us understand the improvement your feature provides. This allows us to directly see the changes in the code and to understand how they affect workflows and performance.

Please see our ✅ Contributing Guide to get started.

[yusiyoh]

Actually, I was asking why you changed the default code for pascal voc ap calculation as in https://github.com/rbgirshick/py-faster-rcnn/blob/781a917b378dbfdedb45b6a56189a31982da1b43/lib/datasets/voc_eval.py#L49-L50 or in https://github.com/facebookresearch/detectron2/blob/11528ce083dc9ff83ee3a8f9086a1ef54d2a402f/detectron2/evaluation/pascal_voc_evaluation.py#L171-L172 and also in https://github.com/open-mmlab/mmdetection/blob/3545915bf14f2acbb5def6355873f7a773149e62/mmdet/core/evaluation/mean_ap.py#L37-L38 . Is there a specific reason for the change that I mentioned?

[glenn-jocher]

@yusiyoh thanks for the links! Yes your 3 (reputable) sources are certainly in alignment on the P and R arrays. Given this I think we should align with your examples. The main reason for the change was that after we started producing P-R curves by default I noticed that this last point connects the curves to the bottom right corner of the plot in all cases, which results in some odd looking curves on occasion, so the current practices terminates the curve shortly after the last data point on the x axis rather than extending it to the x axis maximum.