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.

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

[Feature] Support AnimalKingdom dataset #2139

Merged
merged 49 commits into from May 20, 2023
Merged

[Feature] Support AnimalKingdom dataset #2139

merged 49 commits into from May 20, 2023

Conversation

Dominic23331
Copy link
Contributor

Add AnimalKingdom dataset

@CLAassistant
Copy link

CLAassistant commented Apr 1, 2023

CLA assistant check
All committers have signed the CLA.

@Ben-Louis
Copy link
Collaborator

Thank you so much for your valuable contribution! By the way, have you tried to train any model on this dataset?

@Dominic23331
Copy link
Contributor Author

Not yet. I'll try it in a few days。

@Ben-Louis
Copy link
Collaborator

Not yet. I'll try it in a few days。

Thank you. We shall anticipate the inclusion of the model configuration files. Moreover, it is suggested to incorporate a concise overview of this dataset, together with the file hierarchy, in the docs/en/dataset_zoo/2d_animal_keypoint.md.

@Dominic23331
Copy link
Contributor Author

I upload config files and added explanatory documents

@Ben-Louis
Copy link
Collaborator

I upload config files and added explanatory documents

Thanks. Have you trained models with any of them in order to verify their accuracy?

@Dominic23331
Copy link
Contributor Author

I upload config files and added explanatory documents

Thanks. Have you trained models with any of them in order to verify their accuracy?

I have tried to train several classes, but the results are not very satisfactory. I will consider how to improve it.

@Dominic23331
Copy link
Contributor Author

Due to the busy work ahead, I apologize for the delay in this PR. During testing, I found that 210 epochs were not enough to fit the training set, so I increased the epochs to 300.

@Dominic23331
Copy link
Contributor Author

This is the result of my training
paper mmpose
ALL 66.06 62.72
leave-k-out 39.3 37.74
mammals 61.59 57.56
amphibians 56.74 53.56
reptiles 56.06 50
birds 77.35 76.79
fishes 68.25 64.3
This result is not ideal

@codecov
Copy link

codecov bot commented May 19, 2023

Codecov Report

Patch coverage: 100.00% and project coverage change: +0.02 🎉

Comparison is base (a83e7de) 82.01% compared to head (449b689) 82.03%.

❗ Current head 449b689 differs from pull request most recent head 2666004. Consider uploading reports for the commit 2666004 to get more accurate results

Additional details and impacted files
@@             Coverage Diff             @@
##           dev-1.x    #2139      +/-   ##
===========================================
+ Coverage    82.01%   82.03%   +0.02%     
===========================================
  Files          232      233       +1     
  Lines        13637    13643       +6     
  Branches      2319     2320       +1     
===========================================
+ Hits         11184    11192       +8     
+ Misses        1915     1913       -2     
  Partials       538      538              
Flag Coverage Δ
unittests 82.03% <100.00%> (+0.02%) ⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files Coverage Δ
mmpose/datasets/datasets/animal/__init__.py 100.00% <100.00%> (ø)
.../datasets/datasets/animal/animalkingdom_dataset.py 100.00% <100.00%> (ø)

... and 1 file with indirect coverage changes

☔ View full report in Codecov by Sentry.
📢 Do you have feedback about the report comment? Let us know in this issue.

@Tau-J Tau-J changed the title [feature] add ak dataset [Feature] Support AnimalKingdom dataset May 20, 2023
@Tau-J Tau-J merged commit 65ef573 into open-mmlab:dev-1.x May 20, 2023
15 of 17 checks passed
@Tau-J Tau-J mentioned this pull request May 22, 2023
11 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

10 participants