Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

TrivialAugment image augmentation #413

Closed
innat opened this issue May 5, 2022 · 5 comments
Closed

TrivialAugment image augmentation #413

innat opened this issue May 5, 2022 · 5 comments
Labels
needs-impact-verification Unclear whether or not the feature should be included.

Comments

@innat
Copy link
Contributor

innat commented May 5, 2022

Short Description

From abstract,

... While existing automatic augmentation methods need to trade off simplicity, cost and performance, we present a most simple baseline, TrivialAugment, that outperforms previous methods for almost free. TrivialAugment is parameter-free and only applies a single augmentation to each image. ...

Papers
Paper: TrivialAugment: Tuning-free Yet State-of-the-Art Data Augmentation
https://arxiv.org/abs/2103.10158

Existing Implementations

Other Information

  • (from me, not yet checked thoroughly for a strong feature request.)
@bhack
Copy link
Contributor

bhack commented May 5, 2022

7 citations after ~ 1 year.

It is TA in #294 (comment) (related to #41 (comment))

@LukeWood
Copy link
Contributor

LukeWood commented May 5, 2022

Realistically, I'm not sure we want this given the citation count and usage.

@LukeWood LukeWood added the needs-impact-verification Unclear whether or not the feature should be included. label May 5, 2022
@bhack
Copy link
Contributor

bhack commented May 5, 2022 •
edited
Loading

As we are and we have already invested enough community resources on KLP I suggest to strictly monitor/evaluate the impact of #293 (comment) before accumulating too much other solutions in the repo.

@bhack
Copy link
Contributor

bhack commented May 5, 2022

P.s. I've fixed the reference.

@ianstenbit
Copy link
Contributor

This issue is not currently tied to any near-term objectives of KerasCV, so we are closing it for now. If you think this was done in error, or you are interested in adding this feature yourself, please checkout our README and re-open this issue if it meets the listed criteria.

Thank you! - The KerasCV team

freedomtan pushed a commit to freedomtan/keras-cv that referenced this issue Jul 20, 2023
@mattdangerw
Use tf.range instead of tfnp.arange (keras-team#413)
130f993 
tfnp.arange has trouble with dynamic tensors in compiled functions. The
specific error:

OperatorNotAllowedInGraphError: Using a symbolic `tf.Tensor` as a Python
`bool` is not allowed: AutoGraph did convert this function. This might
indicate you are trying to use an unsupported feature.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
needs-impact-verification Unclear whether or not the feature should be included.
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
4 participants
@bhack @ianstenbit @LukeWood @innat