Devolearn is a Python package that aims to automate the process of collecting metadata from videos/images of the C. elegans embryo with the help of deep learning models. It is specialized for the analysis of 2-D slices of C. elegans embryogenesis.
This project will focus on improving the performance of existing models and training/benchmarking new models on a broad range of data sets from different species relevant to developmental biology. Your goals will be to improve overall performance of the models in terms of accuracy and generalizability, training/adding new models and improve the library’s usability.
There would be 4 key elements in this project:
Improving the current models
Training and adding more useful models
Improving usability
(optional) Interactive online demos.
Improving/adding new models
In order to get good performance from a deep-learning model, we should use techniques like:
k-fold cross validation
Ensembling/stacking
Hyperparameter optimization
Label smoothing
Mixed precision inference
Using more training data for re-training the pre-existing models for better accuracy
Improving usability
As the library grows, we should have a better place to host proper documentation on a static website. We also have to train devolearn’s deep-learning models on commonly available data, the more common the use case, the more helpful it is to the community.
Online demos
Jupyter notebooks, however simple they may seem, are still a bit intimidating to people from non CS backgrounds, so there should be a focus on making interactive and easy to run online demos which showcase both our research and tutorials.
Pre-requisites
Proficiency in python with a good hold of tools like numpy, pandas and PyTorch will be required.
Hey, @malin, I have done some projects with the help of deep learning models. So, have a previous experience with them. I found this project to be much similar with what I have done previously, just need some time to get familiar with code base. Please let me know how to start, till then I will have a look at the code base and other documentries.
I am Vrutik Rabadia, an undergraduate student at the Indian Institute of Information Technology, Surat. I went through all the projects in the incf foundation and found this project in my interest because I have always been interested in deep learning and computer vision projects and it also fits my skill set very well.
I have two months of relevant work experience as a deep learning project intern at the Indian Space Research Organization(ISRO). The purpose of the internship was to implement the You Look only once algorithm for object detection and MPCM fuzzy classifier algorithm for segmentation in the field of remote sensing on large datasets.
I’m looking forward to contributing to this organization and improve my skills.
It would be great if I am able to solve any issue under your guidance.
I am Kshitij Soni, an undergraduate at Indian Institute of Petroleum and Energy Vizag. I have previously worked with IIT Kanpur and published a paper titled " Adverse Drug Reaction Classification using NLP Tools and Deep Learning Methods" . Currently I am working with IIPE Vizag on "Optimization of tool wear and tear by Digital Image Processing.
It will be great opportunity for meif we collaborate to complete this project
Hope all of you are safe and doing well.
I am Devishi, an undergraduate student majoring in electronics. To me this project is really exciting because not only does it show how deep learning techniques prove to be beneficial for biological applications, but also how starting with C. elegans, the same process could potentially be expanded to other organisms thus making it so much easier to study them.
I sincerely wish to get involved in this project. I have gone through the resources given above, but I also have a few doubts. Would it be better to discuss them on Slack or would this platform be okay?
Hello. The best way to find out about the projects and organization is to consult our Onboarding Guide and join the OpenWorm Slack. Once in the Slack, you will want to join the #devoworm, #devolearn, and #gsoc2021 channels. And of course you are welcome to join our weekly meetings, Mondays at 2pm UTC.
