tensorflow / tensorflow/models

Custom loss for Faster R-CNN models

Open
#9,968 0 comments 0 reactions 3 assignees View on GitHub

@pkulzc is already working on this.

Since May 4, 2021.

models:research:odapi type:feature
Dominant language
Python
Stars
77.7k
Forks
44.8k
PR merge metrics
No merged PRs in 30d

Description

Prerequisites

Please answer the following question for yourself before submitting an issue.

  • I checked to make sure that this feature has not been requested already.

1. The entire URL of the file you are using

https://github.com/tensorflow/models/tree/master/research/object_detection/

2. Describe the feature you request

I am using the TensorFlow Object Detection API and I would like to implement a second-stage custom loss for a model that is part of the Faster R-CNN family.
I've read this discussion https://github.com/tensorflow/models/issues/7817 where it is explained how to define a custom loss. However, the problem here is that in the meta architecture FasterRCNNMetaArch a developer could try to "inject" only a custom second-stage classification loss, but the localization loss is hard-coded (it is an instance of WeightedSmoothL1LocalizationLoss).

3. Additional context

--

4. Are you willing to contribute it?

No

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.