End-to-end Malaria Diagnosis and 3D Cell Rendering with Deep Learning

 Vignav Ramesh

arXiv / Full Paper (PDF) / Papers With Code / Website (Malaria Diagnosis Algorithm, 3D Cell Rendering)

Pre-trained models

Model	Pruned (y/n)	Weights
Vanilla CNN	n	v1, v2
VGG-19	n	Download
VGG-19	y	Download

Environment setup

Our models were trained on a single GPU (Tesla P4 GPU provided by Google Colab, 16 GB memory), but the code can be run as is on a CPU (with increased training times). The code is implemented using Python 3, Keras, and TensorFlow v2. To install all required dependencies, run the following:

pip3 install -r requirements.txt

Data

All data is stored in this repository and can be accessed here.

Training

Use the following Colab file to train the model:

Cite

@misc{ramesh2021endtoend,
  title={End-to-end Malaria Diagnosis and 3D Cell Rendering with Deep Learning},
  author={Vignav Ramesh},
  year={2021},
  eprint={2108.04220},
  archivePrefix={arXiv},
  primaryClass={cs.CV}
}