Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/PavlosMelissinos/enet-keras
A keras implementation of ENet (abandoned for the foreseeable future)
https://github.com/PavlosMelissinos/enet-keras
Last synced: about 2 months ago
JSON representation
A keras implementation of ENet (abandoned for the foreseeable future)
- Host: GitHub
- URL: https://github.com/PavlosMelissinos/enet-keras
- Owner: PavlosMelissinos
- License: mit
- Created: 2017-03-30T19:55:12.000Z (over 7 years ago)
- Default Branch: master
- Last Pushed: 2022-12-07T23:36:33.000Z (almost 2 years ago)
- Last Synced: 2024-07-10T11:01:42.100Z (3 months ago)
- Language: Python
- Homepage:
- Size: 282 KB
- Stars: 115
- Watchers: 16
- Forks: 46
- Open Issues: 18
-
Metadata Files:
- Readme: README.md
- License: LICENSE
- Code of conduct: CODE_OF_CONDUCT.md
Awesome Lists containing this project
README
# ENet-keras
[![license](https://img.shields.io/github/license/mashape/apistatus.svg)](https://github.com/PavlosMelissinos/enet-keras/blob/master/LICENSE)
![](https://reposs.herokuapp.com/?path=PavlosMelissinos/enet-keras&style=flat&color=red)
[![Read the Docs](https://img.shields.io/readthedocs/pip.svg)]()This is an implementation of [ENet: A Deep Neural Network Architecture for Real-Time Semantic Segmentation](https://arxiv.org/abs/1606.02147), ported from [ENet-training](https://github.com/e-lab/ENet-training) ([lua-torch](https://github.com/torch/torch7)) to [keras](https://github.com/fchollet/keras).
## Installation
### Get code
```
git clone https://github.com/PavlosMelissinos/enet-keras.git
cd enet-keras
```### Setup environment
#### Dependencies
On poetry: `poetry install`
On Anaconda/miniconda: `conda env create -f environment.yml`
On pip: `pip install -r requirements.txt`
### Set up data/model
`make setup`
The setup script only sets up some directories and converts the model to an appropriate format.
## Usage
### Train on MS-COCO
`make train`
## Remaining tasks
- [ ] Clean up code
- [ ] Remove hardcoded paths
- [ ] Add documentation everywhere
- [ ] Test code
- [ ] Add tests
- [ ] Fix performance (mostly preprocessing bottleneck)
- [ ] Remove unnecessary computations in data preprocessing
- [ ] Index dataset category internals. Dataset categories have fields with one-to-one correspondence like id, category_id, palette, categories. This seems like perfect table structure. Might be too much though.
- [ ] (Optionally) Make data loader multithreaded (no idea how to approach this one, multithreadedness is handled by keras though)
- [ ] Enhance reproducibility/usability
- [x] Upload pretrained model
- [ ] Finalize predict.py
- [x] Test whether it works after latest changes
- [ ] Modify predict.py to load a single image or from a file. There's no point in loading images from the validation set.
- [ ] Fix bugs
- [ ] Investigate reason for bad results, see [#11](https://github.com/PavlosMelissinos/enet-keras/issues/11)
- [ ] Fix MSCOCOReduced, [also see #9](https://github.com/PavlosMelissinos/enet-keras/issues/9)
- [ ] ?????