Awesome
<div align="center">Accelerated DL R&D
</div>PyTorch framework for Deep Learning research and development. It was developed with a focus on reproducibility, fast experimentation and code/ideas reusing. Being able to research/develop something new, rather than write another regular train loop. <br/> Break the cycle - use the Catalyst!
Project manifest. Part of PyTorch Ecosystem. Part of Catalyst Ecosystem:
- Alchemy - Experiments logging & visualization
- Catalyst - Accelerated Deep Learning Research and Development
- Reaction - Convenient Deep Learning models serving
Catalyst.Detection
Note: this repo uses advanced Catalyst Config API and could be a bit out-of-day right now. Use Catalyst's minimal examples section for a starting point and up-to-day use cases, please.
Based on Objects as points article by Xingyi Zhou, Dequan Wang, Philipp Krähenbühl
Training in your dataset
-
Install requirements
pip install -r requirements.txt
-
Copy all images to one directory or two different directories for train and validation.
-
Create
markup_train.json
as json file in MSCOCO format usingCOCODetectionFactory
fromdata_preparation.py
. This class may be copied to your dataset generator. See documentation in code comments. If your dataset are already in this format, go to next step. -
Specify perameters and in
config/centernet_detection_config.yml
. -
Run catalyst
catalyst-dl run --config=./configs/centernet_detection_config.yml
-
When you change dataset, you must delete cache files
markup_*.json.cache
because this files contain preprocessed bounding boxes info.