https://github.com/ankandrew/yolox-models
Object Detection models trained for different tasks (i.e. face, person, etc.) with different models (i.e. nano, tiny, etc.)
https://github.com/ankandrew/yolox-models
face-detection object-detection object-detections-models yolox yolox-models
Last synced: about 2 months ago
JSON representation
Object Detection models trained for different tasks (i.e. face, person, etc.) with different models (i.e. nano, tiny, etc.)
- Host: GitHub
- URL: https://github.com/ankandrew/yolox-models
- Owner: ankandrew
- License: mit
- Created: 2022-01-15T15:26:47.000Z (over 3 years ago)
- Default Branch: main
- Last Pushed: 2024-05-16T15:58:37.000Z (11 months ago)
- Last Synced: 2025-01-03T21:31:38.853Z (4 months ago)
- Topics: face-detection, object-detection, object-detections-models, yolox, yolox-models
- Language: Jupyter Notebook
- Homepage:
- Size: 15.6 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
## YOLOX Trained Models
Have you ever wanted a **fast**/**accurate** object detection model, but didn't have **time**/**resources** to train it?
### Intro
This repo aims to provide a variety **trained models** on **different objects**, which you can use directly
with [YOLOX](https://github.com/Megvii-BaseDetection/YOLOX) framework.
We also train models on different views of the datasets that have bigger/lower **object area**. For example, if you need
a face detector that will be used on mobile front cameras, there may be no need to detect faces that have very
**small area** (very far objects). We can take advantage of this and use a lower input resolution, so inference is
very **fast**.
### Contents
- [Request Model](#request-model)
- [Models](#models)
* [Face](#face)
### Models HUB 🚀
Models are mainly trained with Open [Images Dataset V6](https://storage.googleapis.com/openimages/web/index.html), which
has 600 classes. To re-create the dataset used in the trained models, refer to [dataset.ipynb](dataset.ipynb).
### Models
#### Face
| Model | Activation | Input Resolution | mAPtest
0.5:0.95 | Area | Weights | Experiment |
|-------|------------|------------------|--------------------------------|-------|----------------------------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------------|
| nano | silu | 128x128 | 74.20 | > 10% | [github](https://github.com/ankandrew/yolox-models/releases/download/v1.0.0/faces_10p_area_nano_128_silu.pth) | [github](https://github.com/ankandrew/yolox-models/releases/download/v1.0.0/nano_128_silu.py) |
| nano | leaky relu | 128x128 | 73.68 | > 10% | [github](https://github.com/ankandrew/yolox-models/releases/download/v1.0.0/faces_10p_area_nano_128_lrelu.pth) | [github](https://github.com/ankandrew/yolox-models/releases/download/v1.0.0/nano_128_lrelu.py) |
| nano | silu | 160x160 | 72.72 | > 5% | [github](https://github.com/ankandrew/yolox-models/releases/download/v1.0.0/faces_5p_area_nano_160_silu.pth) | [github](https://github.com/ankandrew/yolox-models/releases/download/v1.0.0/nano_160_silu.py) |
| nano | leaky relu | 160x160 | 71.90 | > 5% | [github](https://github.com/ankandrew/yolox-models/releases/download/v1.0.0/faces_5p_area_nano_160_lrelu.pth) | [github](https://github.com/ankandrew/yolox-models/releases/download/v1.0.0/nano_160_lrelu.py) |
| nano | silu | 192x192 | 66.97 | > 1% | [github](https://github.com/ankandrew/yolox-models/releases/download/v1.0.0/faces_1p_area_nano_192_silu.pth) | [github](https://github.com/ankandrew/yolox-models/releases/download/v1.0.0/nano_192_silu.py) |
| nano | leaky relu | 192x192 | 66.21 | > 1% | [github](https://github.com/ankandrew/yolox-models/releases/download/v1.0.0/faces_1p_area_nano_192_lrelu.pth) | [github](https://github.com/ankandrew/yolox-models/releases/download/v1.0.0/nano_192_lrelu.py) |
_Note: You can try using any model with slightly higher/lower input resolution and will also work fine._
### TODO
- [x] Make script to reproduce datasets
### Reference
```latex
@article{yolox2021,
title={YOLOX: Exceeding YOLO Series in 2021},
author={Ge, Zheng and Liu, Songtao and Wang, Feng and Li, Zeming and Sun, Jian},
journal={arXiv preprint arXiv:2107.08430},
year={2021}
}
```