Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

https://github.com/fizyr/keras-retinanet

Keras implementation of RetinaNet object detection.
https://github.com/fizyr/keras-retinanet

Last synced: 4 months ago
JSON representation

Keras implementation of RetinaNet object detection.

Host: GitHub
URL: https://github.com/fizyr/keras-retinanet
Owner: fizyr
License: apache-2.0
Created: 2017-08-14T09:14:29.000Z (almost 7 years ago)
Default Branch: main
Last Pushed: 2023-03-16T14:51:24.000Z (over 1 year ago)
Last Synced: 2024-02-28T01:04:53.895Z (4 months ago)
Language: Python
Size: 13.2 MB
Stars: 4,352
Watchers: 127
Forks: 1,975
Open Issues: 29
Metadata Files:
- Readme: README.md
- License: LICENSE

Lists

my-awesome-stars - fizyr/keras-retinanet - Keras implementation of RetinaNet object detection. (Python)
awesome_quantmetry - retinanet

README

# Keras RetinaNet [![Build Status](https://travis-ci.org/fizyr/keras-retinanet.svg?branch=master)](https://travis-ci.org/fizyr/keras-retinanet) [![DOI](https://zenodo.org/badge/100249425.svg)](https://zenodo.org/badge/latestdoi/100249425)

Keras implementation of RetinaNet object detection as described in [Focal Loss for Dense Object Detection](https://arxiv.org/abs/1708.02002)
by Tsung-Yi Lin, Priya Goyal, Ross Girshick, Kaiming He and Piotr Dollár.

## :warning: Deprecated

This repository is deprecated in favor of the [torchvision](https://github.com/pytorch/vision/) module.
This project should work with keras 2.4 and tensorflow 2.3.0, newer versions might break support.
For more information, check [here](https://github.com/fizyr/keras-retinanet/issues/1471#issuecomment-704187205).

## Installation

1) Clone this repository.
2) In the repository, execute `pip install . --user`.
Note that due to inconsistencies with how `tensorflow` should be installed,
this package does not define a dependency on `tensorflow` as it will try to install that (which at least on Arch Linux results in an incorrect installation).
Please make sure `tensorflow` is installed as per your systems requirements.
3) Alternatively, you can run the code directly from the cloned repository, however you need to run `python setup.py build_ext --inplace` to compile Cython code first.
4) Optionally, install `pycocotools` if you want to train / test on the MS COCO dataset by running `pip install --user git+https://github.com/cocodataset/cocoapi.git#subdirectory=PythonAPI`.

## Testing
An example of testing the network can be seen in [this Notebook](https://github.com/delftrobotics/keras-retinanet/blob/master/examples/ResNet50RetinaNet.ipynb).
In general, inference of the network works as follows:
```python
boxes, scores, labels = model.predict_on_batch(inputs)
```

Where `boxes` are shaped `(None, None, 4)` (for `(x1, y1, x2, y2)`), scores is shaped `(None, None)` (classification score) and labels is shaped `(None, None)` (label corresponding to the score). In all three outputs, the first dimension represents the shape and the second dimension indexes the list of detections.

Loading models can be done in the following manner:
```python
from keras_retinanet.models import load_model
model = load_model('/path/to/model.h5', backbone_name='resnet50')
```

Execution time on NVIDIA Pascal Titan X is roughly 75msec for an image of shape `1000x800x3`.

### Converting a training model to inference model
The training procedure of `keras-retinanet` works with *training models*. These are stripped down versions compared to the *inference model* and only contains the layers necessary for training (regression and classification values). If you wish to do inference on a model (perform object detection on an image), you need to convert the trained model to an inference model. This is done as follows:

```shell
# Running directly from the repository:
keras_retinanet/bin/convert_model.py /path/to/training/model.h5 /path/to/save/inference/model.h5

# Using the installed script:
retinanet-convert-model /path/to/training/model.h5 /path/to/save/inference/model.h5
```

Most scripts (like `retinanet-evaluate`) also support converting on the fly, using the `--convert-model` argument.

## Training
`keras-retinanet` can be trained using [this](https://github.com/fizyr/keras-retinanet/blob/master/keras_retinanet/bin/train.py) script.
Note that the train script uses relative imports since it is inside the `keras_retinanet` package.
If you want to adjust the script for your own use outside of this repository,
you will need to switch it to use absolute imports.

If you installed `keras-retinanet` correctly, the train script will be installed as `retinanet-train`.
However, if you make local modifications to the `keras-retinanet` repository, you should run the script directly from the repository.
That will ensure that your local changes will be used by the train script.

The default backbone is `resnet50`. You can change this using the `--backbone=xxx` argument in the running script.
`xxx` can be one of the backbones in resnet models (`resnet50`, `resnet101`, `resnet152`), mobilenet models (`mobilenet128_1.0`, `mobilenet128_0.75`, `mobilenet160_1.0`, etc), densenet models or vgg models. The different options are defined by each model in their corresponding python scripts (`resnet.py`, `mobilenet.py`, etc).

Trained models can't be used directly for inference. To convert a trained model to an inference model, check [here](https://github.com/fizyr/keras-retinanet#converting-a-training-model-to-inference-model).

### Usage
For training on [Pascal VOC](http://host.robots.ox.ac.uk/pascal/VOC/), run:
```shell
# Running directly from the repository:
keras_retinanet/bin/train.py pascal /path/to/VOCdevkit/VOC2007

# Using the installed script:
retinanet-train pascal /path/to/VOCdevkit/VOC2007
```

For training on [MS COCO](http://cocodataset.org/#home), run:
```shell
# Running directly from the repository:
keras_retinanet/bin/train.py coco /path/to/MS/COCO

# Using the installed script:
retinanet-train coco /path/to/MS/COCO
```

For training on Open Images Dataset [OID](https://storage.googleapis.com/openimages/web/index.html)
or taking place to the [OID challenges](https://storage.googleapis.com/openimages/web/challenge.html), run:
```shell
# Running directly from the repository:
keras_retinanet/bin/train.py oid /path/to/OID

# Using the installed script:
retinanet-train oid /path/to/OID

# You can also specify a list of labels if you want to train on a subset
# by adding the argument 'labels_filter':
keras_retinanet/bin/train.py oid /path/to/OID --labels-filter=Helmet,Tree

# You can also specify a parent label if you want to train on a branch
# from the semantic hierarchical tree (i.e a parent and all children)
(https://storage.googleapis.com/openimages/challenge_2018/bbox_labels_500_hierarchy_visualizer/circle.html)
# by adding the argument 'parent-label':
keras_retinanet/bin/train.py oid /path/to/OID --parent-label=Boat
```

For training on [KITTI](http://www.cvlibs.net/datasets/kitti/eval_object.php), run:
```shell
# Running directly from the repository:
keras_retinanet/bin/train.py kitti /path/to/KITTI

# Using the installed script:
retinanet-train kitti /path/to/KITTI

If you want to prepare the dataset you can use the following script:
https://github.com/NVIDIA/DIGITS/blob/master/examples/object-detection/prepare_kitti_data.py
```

For training on a [custom dataset], a CSV file can be used as a way to pass the data.
See below for more details on the format of these CSV files.
To train using your CSV, run:
```shell
# Running directly from the repository:
keras_retinanet/bin/train.py csv /path/to/csv/file/containing/annotations /path/to/csv/file/containing/classes

# Using the installed script:
retinanet-train csv /path/to/csv/file/containing/annotations /path/to/csv/file/containing/classes
```

In general, the steps to train on your own datasets are:
1) Create a model by calling for instance `keras_retinanet.models.backbone('resnet50').retinanet(num_classes=80)` and compile it.
Empirically, the following compile arguments have been found to work well:
```python
model.compile(
loss={
'regression' : keras_retinanet.losses.smooth_l1(),
'classification': keras_retinanet.losses.focal()
},
optimizer=keras.optimizers.Adam(lr=1e-5, clipnorm=0.001)
)
```
2) Create generators for training and testing data (an example is show in [`keras_retinanet.preprocessing.pascal_voc.PascalVocGenerator`](https://github.com/fizyr/keras-retinanet/blob/master/keras_retinanet/preprocessing/pascal_voc.py)).
3) Use `model.fit_generator` to start training.

## Pretrained models

All models can be downloaded from the [releases page](https://github.com/fizyr/keras-retinanet/releases).

### MS COCO

Results using the `cocoapi` are shown below (note: according to the paper, this configuration should achieve a mAP of 0.357).

### Open Images Dataset
There are 3 RetinaNet models based on ResNet50, ResNet101 and ResNet152 trained on all [500 classes](https://github.com/ZFTurbo/Keras-RetinaNet-for-Open-Images-Challenge-2018/blob/master/a00_utils_and_constants.py#L130) of the Open Images Dataset (thanks to @ZFTurbo).

| Backbone | Image Size (px) | Small validation mAP | LB (Public) |
| --------- | --------------- | -------------------- | ----------- |
| ResNet50 | 768 - 1024 | 0.4594 | 0.4223 |
| ResNet101 | 768 - 1024 | 0.4986 | 0.4520 |
| ResNet152 | 600 - 800 | 0.4991 | 0.4651 |

For more information, check [@ZFTurbo's](https://github.com/ZFTurbo/Keras-RetinaNet-for-Open-Images-Challenge-2018) repository.

## CSV datasets
The `CSVGenerator` provides an easy way to define your own datasets.
It uses two CSV files: one file containing annotations and one file containing a class name to ID mapping.

### Annotations format
The CSV file with annotations should contain one annotation per line.
Images with multiple bounding boxes should use one row per bounding box.
Note that indexing for pixel values starts at 0.
The expected format of each line is:
```
path/to/image.jpg,x1,y1,x2,y2,class_name
```
By default the CSV generator will look for images relative to the directory of the annotations file.

Some images may not contain any labeled objects.
To add these images to the dataset as negative examples,
add an annotation where `x1`, `y1`, `x2`, `y2` and `class_name` are all empty:
```
path/to/image.jpg,,,,,
```

A full example:
```
/data/imgs/img_001.jpg,837,346,981,456,cow
/data/imgs/img_002.jpg,215,312,279,391,cat
/data/imgs/img_002.jpg,22,5,89,84,bird
/data/imgs/img_003.jpg,,,,,
```

This defines a dataset with 3 images.
`img_001.jpg` contains a cow.
`img_002.jpg` contains a cat and a bird.
`img_003.jpg` contains no interesting objects/animals.

### Class mapping format
The class name to ID mapping file should contain one mapping per line.
Each line should use the following format:
```
class_name,id
```

Indexing for classes starts at 0.
Do not include a background class as it is implicit.

For example:
```
cow,0
cat,1
bird,2
```

## Anchor optimization

In some cases, the default anchor configuration is not suitable for detecting objects in your dataset, for example, if your objects are smaller than the 32x32px (size of the smallest anchors). In this case, it might be suitable to modify the anchor configuration, this can be done automatically by following the steps in the [anchor-optimization](https://github.com/martinzlocha/anchor-optimization/) repository. To use the generated configuration check [here](https://github.com/fizyr/keras-retinanet-test-data/blob/master/config/config.ini) for an example config file and then pass it to `train.py` using the `--config` parameter.

## Debugging
Creating your own dataset does not always work out of the box. There is a [`debug.py`](https://github.com/fizyr/keras-retinanet/blob/master/keras_retinanet/bin/debug.py) tool to help find the most common mistakes.

Particularly helpful is the `--annotations` flag which displays your annotations on the images from your dataset. Annotations are colored in green when there are anchors available and colored in red when there are no anchors available. If an annotation doesn't have anchors available, it means it won't contribute to training. It is normal for a small amount of annotations to show up in red, but if most or all annotations are red there is cause for concern. The most common issues are that the annotations are too small or too oddly shaped (stretched out).

## Results

### MS COCO

## Status
Example output images using `keras-retinanet` are shown below.

Example result of RetinaNet on MS COCO

### Projects using keras-retinanet
* [Improving Apple Detection and Counting Using RetinaNet](https://github.com/nikostsagk/Apple-detection). This work aims to investigate the apple detection problem through the deployment of the Keras RetinaNet.
* [Improving RetinaNet for CT Lesion Detection with Dense Masks from Weak RECIST Labels](https://arxiv.org/abs/1906.02283). Research project for detecting lesions in CT using keras-retinanet.
* [NudeNet](https://github.com/bedapudi6788/NudeNet). Project that focuses on detecting and censoring of nudity.
* [Individual tree-crown detection in RGB imagery using self-supervised deep learning neural networks](https://www.biorxiv.org/content/10.1101/532952v1). Research project focused on improving the performance of remotely sensed tree surveys.
* [ESRI Object Detection Challenge 2019](https://github.com/kunwar31/ESRI_Object_Detection). Winning implementation of the ESRI Object Detection Challenge 2019.
* [Lunar Rockfall Detector Project](https://ieeexplore.ieee.org/document/8587120). The aim of this project is to [map lunar rockfalls on a global scale](https://www.nature.com/articles/s41467-020-16653-3) using the available > 2 million satellite images.
* [Mars Rockfall Detector Project](https://ieeexplore.ieee.org/document/9103997). The aim of this project is to map rockfalls on Mars.
* [NATO Innovation Challenge](https://medium.com/data-from-the-trenches/object-detection-with-deep-learning-on-aerial-imagery-2465078db8a9). The winning team of the NATO Innovation Challenge used keras-retinanet to detect cars in aerial images ([COWC dataset](https://gdo152.llnl.gov/cowc/)).
* [Microsoft Research for Horovod on Azure](https://blogs.technet.microsoft.com/machinelearning/2018/06/20/how-to-do-distributed-deep-learning-for-object-detection-using-horovod-on-azure/). A research project by Microsoft, using keras-retinanet to distribute training over multiple GPUs using Horovod on Azure.
* [Anno-Mage](https://virajmavani.github.io/saiat/). A tool that helps you annotate images, using input from the keras-retinanet COCO model as suggestions.
* [Telenav.AI](https://github.com/Telenav/Telenav.AI/tree/master/retinanet). For the detection of traffic signs using keras-retinanet.
* [Towards Deep Placental Histology Phenotyping](https://github.com/Nellaker-group/TowardsDeepPhenotyping). This research project uses keras-retinanet for analysing the placenta at a cellular level.
* [4k video example](https://www.youtube.com/watch?v=KYueHEMGRos). This demo shows the use of keras-retinanet on a 4k input video.
* [boring-detector](https://github.com/lexfridman/boring-detector). I suppose not all projects need to solve life's biggest questions. This project detects the "The Boring Company" hats in videos.
* [comet.ml](https://towardsdatascience.com/how-i-monitor-and-track-my-machine-learning-experiments-from-anywhere-described-in-13-tweets-ec3d0870af99). Using keras-retinanet in combination with [comet.ml](https://comet.ml) to interactively inspect and compare experiments.
* [Weights and Biases](https://app.wandb.ai/syllogismos/keras-retinanet/reports?view=carey%2FObject%20Detection%20with%20RetinaNet). Trained keras-retinanet on coco dataset from beginning on resnet50 and resnet101 backends.
* [Google Open Images Challenge 2018 15th place solution](https://github.com/ZFTurbo/Keras-RetinaNet-for-Open-Images-Challenge-2018). Pretrained weights for keras-retinanet based on ResNet50, ResNet101 and ResNet152 trained on open images dataset.
* [poke.AI](https://github.com/Raghav-B/poke.AI). An experimental AI that attempts to master the 3rd Generation Pokemon games. Using keras-retinanet for in-game mapping and localization.
* [retinanetjs](https://github.com/faustomorales/retinanetjs). A wrapper to run RetinaNet inference in the browser / Node.js. You can also take a look at the [example app](https://faustomorales.github.io/retinanetjs-example-app/).
* [CRFNet](https://github.com/TUMFTM/CameraRadarFusionNet). This network fuses radar and camera data to perform object detection for autonomous driving applications.
* [LogoDet](https://github.com/notAI-tech/LogoDet). Project for detecting company logos in images.
* [AIR: Aerial Inspection RetinaNet](https://github.com/Accenture/AIR). A deep learning solution for supporting land search and rescue missions with drones.

If you have a project based on `keras-retinanet` and would like to have it published here, shoot me a message on Slack.

### Notes
* This repository requires Tensorflow 2.3.0 or higher.
* This repository is [tested](https://github.com/fizyr/keras-retinanet/blob/master/.travis.yml) using OpenCV 3.4.
* This repository is [tested](https://github.com/fizyr/keras-retinanet/blob/master/.travis.yml) using Python 2.7 and 3.6.

Contributions to this project are welcome.

### Discussions
Feel free to join the `#keras-retinanet` [Keras Slack](https://keras-slack-autojoin.herokuapp.com/) channel for discussions and questions.

## FAQ
* **I get the warning `UserWarning: No training configuration found in save file: the model was not compiled. Compile it manually.`, should I be worried?** This warning can safely be ignored during inference.
* **I get the error `ValueError: not enough values to unpack (expected 3, got 2)` during inference, what to do?**. This is because you are using a train model to do inference. See https://github.com/fizyr/keras-retinanet#converting-a-training-model-to-inference-model for more information.
* **How do I do transfer learning?** The easiest solution is to use the `--weights` argument when training. Keras will load models, even if the number of classes don't match (it will simply skip loading of weights when there is a mismatch). Run for example `retinanet-train --weights snapshots/some_coco_model.h5 pascal /path/to/pascal` to transfer weights from a COCO model to a PascalVOC training session. If your dataset is small, you can also use the `--freeze-backbone` argument to freeze the backbone layers.
* **How do I change the number / shape of the anchors?** The train tool allows to pass a configuration file, where the anchor parameters can be adjusted. Check [here](https://github.com/fizyr/keras-retinanet-test-data/blob/master/config/config.ini) for an example config file.
* **I get a loss of `0`, what is going on?** This mostly happens when none of the anchors "fit" on your objects, because they are most likely too small or elongated. You can verify this using the [debug](https://github.com/fizyr/keras-retinanet#debugging) tool.
* **I have an older model, can I use it after an update of keras-retinanet?** This depends on what has changed. If it is a change that doesn't affect the weights then you can "update" models by creating a new retinanet model, loading your old weights using `model.load_weights(weights_path, by_name=True)` and saving this model. If the change has been too significant, you should retrain your model (you can try to load in the weights from your old model when starting training, this might be a better starting position than ImageNet).
* **I get the error `ModuleNotFoundError: No module named 'keras_retinanet.utils.compute_overlap'`, how do I fix this?** Most likely you are running the code from the cloned repository. This is fine, but you need to compile some extensions for this to work (`python setup.py build_ext --inplace`).
* **How do I train on my own dataset?** The steps to train on your dataset are roughly as follows:
* 1. Prepare your dataset in the CSV format (a training and validation split is advised).
* 2. Check that your dataset is correct using `retinanet-debug`.
* 3. Train retinanet, preferably using the pretrained COCO weights (this gives a **far** better starting point, making training much quicker and accurate). You can optionally perform evaluation of your validation set during training to keep track of how well it performs (advised).
* 4. Convert your training model to an inference model.
* 5. Evaluate your inference model on your test or validation set.
* 6. Profit!