FineGAN - Partial Implementation TF 2.3 (Community)

[Vishal-V]

Description

📝 Implement Hierarchical Disentanglement for Fine-Grained Object Generation - FineGAN.

• State-of-the-art for CUB 200, CUB 128x128, Stanford Cars and Stanford Dogs datasets.
• Hierarchical Deep Generative Model (Several moving parts trained end-to-end).
• Reproduce the paper results and benchmark it on the 3 datasets.
• Publish documentation and prepare Colab demo.
• Publish pre-trained SavedModel files at TensorFlow Hub.
• Completed implementation and merge solves [Help wanted] FineGAN: Unsupervised Hierarchical Disentanglement for Fine-Grained Object Generation and Discovery #8705

Type of change

• A new research paper code implementation

Key Features

• TensorFlow 2.3.0
• Inference example (Demo)
• Eager mode training with tf.GradientTape [If Required]
• [Default] Graph mode training with model.train_on_batch
• Functional model with tf.keras.layers.Layer
• Subclassing model with tf.keras.Model
• Input pipeline using tf.data and tfds
• Tensorflow Serving
• Vectorized transformations
• GPU accelerated
• Clean implementation
• Following the best practices
• Apache 2.0 License

Tests

📝 [WIP] Code Reviews and Tests Pending.

• [WIP] Provide instructions so we can reproduce.
• [WIP] Please also list any relevant details for your test configuration.

Test Configuration:

• Config .yml files and Config class included.

Checklist

• I have signed the Contributor License Agreement.
• I have read guidelines for pull request.
• My code follows the coding guidelines.
• I have performed a self code review of my own code.
• I have commented my code, particularly in hard-to-understand areas.
• I have made corresponding changes to the documentation.
• My changes generate no new warnings.
• I have added tests that prove my fix is effective or that my feature works.

[review-notebook-app]

Check out this pull request on 

Review Jupyter notebook visual diffs & provide feedback on notebooks.

---

Powered by ReviewNB

[jaeyounkim]

The community directory is for providing a curated list of community models. Please host the FineGan implementation using your own GitHub repository, and then we can review your implementation before adding it to the community directory.