Explore Projects

<p>The main objective of the project is to build a consistent API leveraging MNE-Python to perform connectivity analysis of neural time-series data, such as MEG, EEG and iEEG. The result would be “mne-connectivity” a Python3.6+ software package that allows scientists to explore connectivity based analysis using published algorithms with few barriers because it leverages the widely used open-source platform of MNE-Python.</p>

The goal of this project is to create a set of high-level data preprocessing APIs that allow Hub users to improve overall dataset quality with minimal parameters. A machine-learning algorithm may perform differently on datasets with different characteristics (e.g., it might perform better on a dataset with continuous attributes rather than with categorical attributes). Taking into account all the possible pre-processing strategies, there exists an extremely large number of options and non-experienced users become overwhelmed. A lot of research has been done for providing help and an overview of the different steps of data analysis. The focus, however, has usually been on the model training step, and data pre-processing has generally been overlooked. This problem can be addressed by an automated approach, leveraging ideas from multiple research papers.

Rustup is the official toolchain installer and version manager for the Rust programming language. The current implementation of Rustup processes operations sequentially, downloading components one at a time and performing extraction and installation in order. This approach does not fully leverage the capabilities of modern hardware, particularly on systems with multi-core processors and high-bandwidth connections. A potential solution is to introduce concurrency into the implementation, enabling parallel processing of these tasks to better utilize system resources, leading to a more performant Rustup.

<p><a href="https://knative.dev/" target="_blank">Knative</a> is an open-source community project that extends Kubernetes with modules for deploying, operating, and maintaining serverless, cloud-native applications. The serverless cloud computing model has the potential to improve developer efficiency while decreasing operating costs.The project idea is to develop Charmil, an open-source Golang-based platform that will serve as a package manager which assists in the discovery and installation of plugins from a community-curated plugin index. These plugins can then be embedded into any host CLI with additional features such as new version detection and smooth upgrades.Through Charmil, each CLI will be able to install other CLIs as extensions, thereby organizing them in a hierarchy. This functionality of Charmil will later be used to add new features on top of the Knative CLI.</p>

<p>Smart HTTP/2-based protocol for Infinispan.</p> <p>The Infinispan Server distribution contains a server module that implements Infinispan’s custom binary protocol called Hot Rod. The protocol was designed to enable faster client/server interactions compared to other existing text based protocols and to allow clients to make more intelligent decisions with regards to load balancing, failover and even data location operations.</p> <p>In 2015 the 2nd version of HTTP protocol (simply HTTP/2) was finally introduced. HTTP/2 provides a lot of features that are also implemented in Hot Rod protocol. For example, using the HTTP/2 will be more preferable than Hot Rod because the size of the requests and responses will be smaller.</p> <p>Expected results of this project are:</p> <ol> <li>Modify current REST Server endpoint for Infinispan to handle HTTP/2 protocol. Modification will be based on implementing features from Hot Rod protocol to HTTP/2.</li> <li>Create HTTP client on Java.</li> </ol>

Enigma’s future plan includes becoming more independent by shifting focus away from Game Maker. As Tiled offers better room editing features than Game Maker, adding Tiled compatibility to Enigma can help users take advantage of Tiled room editor along with Enigma’s development environment. This project also expects some research to know the intermediate representation of Tiled in much more detail. TMX is a Tiled specific room format and EGM is an Enigma specific project format. Most of the features supported by EGM format are based on Game Maker’s format GMX. The way this project aims to achieve Tiled compatibility is twofold. First part consists of adding a TMX importer to Enigma while the second part consists of adding an EGM importer to Tiled. Additional tasks involve analyzing GMX 1.4 to understand various protocol buffer fields, adding Gtests in libEGM to check correctness of deserializer efficiently.

<p>The aim of this project is to integrate an automatic AI-based scheme into the gprMax environment such that it provides real-time FDTD solutions to the user. A Deep Learning-based algorithm can be used to provide real-time solutions based on the user's inputs, thus speeding up the computational process. An AI-based real-time EM solver will be orders of magnitude faster than conventional FDTD and would also alleviate the need for having heavy computational resources at the user’s end.</p> <p>In practice, the goal is to develop user-friendly tools with which the user will be able to effectively parametrise the investigated problem and define the expected range of the parameters. Big data will be generated in an automatic manner to be subsequently used for training a deep learning scheme to predict the electromagnetic (EM) response subject to the parameters of the models.</p>

When utilising numerical simulations, particularly with large systems, it can be invaluable to have strong visualization tools to help gain a better intuition and understanding of the system dynamics. It is my aim to implement a fast, lightweight and feature-rich visualization method for PyElastica for during run-time and post-run-time visualization of simulations, that can provide real value to the PyElastica community.

<p>I will be working on adding a new functionality to FURY which shall enable users to visualise various types of proteins via different representations like Richardson aka Ribbon diagrams and molecular surface diagrams. As a part of my stretch goals, I’d like to expand protein representations via other representations including -</p> <ul> <li>Stick</li> <li>Ball and stick</li> <li>Wire</li> <li>Pipes and Planks</li> <li>Sphere</li> </ul>

<p>I will be working on a couple of UI components for FURY’s UI module. These 2D and 3D components will be sci-fi like as seen in <a href="https://www.youtube.com/watch?v=b0ve2nHEVWw" target="_blank">this</a> scene from the movie “Guardians of The Galaxy”.</p> <p>My main objective would be to develop these UI components with their respective test and tutorials such that it adds on to the UI module of FURY and doesn’t hinder existing functionalities/performance.</p>

<p>The aim of this project is to improve data visualization in MNE when working in the Jupyter Notebook. This will be achieved by implementing a package for interactive 3D visualization of human brain activity inside the notebooks. A working alternative to mayavi and PySurfer built on top of ipyvolume and ipywidgets will be developed. Ultimately, I will add ipyvolume as a backend to the MNE package. Export of data visualization to a variety of file formats will be added.</p>

<p>Lunatik is a programming environment based on the Lua language for kernel scripting. By allowing kernel scripting, a user may create scripts in Lua to customize their kernel to suit their needs. Modern day kernels, such as Linux, work on a highly concurrent environment, and therefore must use robust synchronization APIs to ensure data consistency, with each API having their own use cases.</p> <p>The Read-copy-update (RCU) is one of those APIs, made specifically for scenarios where data reading is much more common than writing. RCU allows concurrent readers to access protected data in a non-blocking way even during updates or removals. RCU, as with the rest of the Linux Kernel, is written in C.</p> <p>The project idea is a binding of RCU for use in Lunatik, thus allowing synchronization of Lua data via RCU.</p>

<p>The project highlights use-cases that explain the Machine Learning workflow of DFFML API, through Jupyter notebook examples. The project also adds multi-output model support to DFFML.</p>

This project aims to implement a Near-to-Far Field Transformation (NFFT) feature in gprMax. This will enable users to compute the far-field radiation patterns and radar crosssections(RCS) from the near-field data. It will expand gprMax’s capabilities for antenna design, scattering analysis, etc. I propose to implement the Near-to-Far Field Transformation (NFFT) module and its integration at several key points in the existing workflow: - Adding the NFFT command - Module for Field Sampling during FDTD - Post-Processing & Far-Field Calculation - Integration with the API - Validation and Testing - Documentation and Review

<p>The idea of this project is to use Haskell as another way of reasoning about the typing rules of Typed Lua. Project involves implementing parser and typechecker for Typed Lua Core - Typed Lua dialect which does not contain any syntactic sugar what simplifies reasoning about typing rules.</p>

<p>Continue the implementation of the course with Runestone components. Currently, some exercises are not working due to the Python modules they use. I have in mind replacing them with other modules. The expected outcome is to have the existing lectures implemented in Runestone fully functional with various exercises with automatic grading, have created new lectures teaching other modules, and have implemented other lectures using Brython.</p>

<p>JBoss Web Servers consists of Apache Tomcat 7, Apache Tomcat 8 and Apache Tomcat 8.5 servers. JBoss Core Services consists of Apache HTTP server and its dependencies. The project’s goal is to develop a tool that will scan and parse the upstream security advisories of Apache Tomcat and Apache HTTP Server along with their dependencies. This tool’s purpose is to populate Bugzilla by tracing down any new public CVE on every JWS and JBCS related upstream project. The tool will be built in Python3, scripts will be fired via Jenkins job and scripts will be running as a shell script.</p>

<p>Provide a new mason package implementing various parallel safe sequential data structures.</p>

<p>Few people realise it, but even the modest computers today, including mobile phones, have powerful GPUs. And these GPUs can be used serially and in parallel to CPUs, potentially delivering awesome performance.</p> <p>In this project I want to make it possible to combine the beauty of Ruby with transparent GPU processing so that software developers can easily use that power when available, and farm out computations transparently to GPU and CPU. This will work both on client computers and on servers that make use of TESLA's and Intel Xeon Phi solutions.</p>

<p>I propose to work on a tool that, given JVM bytecode, decides for each method, whether it performs an effectful computation (including IO, global variable mutation etc.), or it is referentially transparent, that is it yields the same result everytime it is called with same arguments. Such functions are called 'pure functions'. The tool is also supposed to provide information about whether the return type can be the special <code>null</code> value, requiring a <code>Maybe a</code> during Eta-interop. Automated analysis for such data can simplify the development process, by enabling an ffi-generator to generate bindings for Java , ideally without requiring any manual input from the developer. Since the project is mostly exploratory, the analysis is intended to be conservative, but never wrong, in the sense that it should not recommend a return type of <code>a</code> for a function that may return <code>null</code>, but may recommend <code>Maybe a</code> for a function that never returns <code>null</code>. But the aim is to, of course, minimize the number of such cases.</p>

<p>DIPY currently uses non-local means approach for denoising MR data, in which we need to estimate the noise variance of the signal which is often a bit troublesome, also it does not make the full use of the directional information in dMRI datasets. This project proposes to use a more robust and efficient method for automatically denoising diffusion MRI and structural MR datasets, using Local-PCA. Along with an accurate implementation of L-PCA and its adaptation to Rician noise, the project will also aim at optimising the implementation of L-PCA using Cython. After implementing L-PCA, a method for robust brain extraction needs to be developed. DIPY’s median OTSU based implementation is known not work so well with non-echo planar diffusion imaging (non-EPI dMRI) data. There are few possible ways to improve this, one is to generate labels and have weighted median OTSU. Another idea involves using some version of patch based segmentation using image library constructed from previously annotated images and using it as a reference for the subsequent extraction.</p>

<p>My proposal is about going through the library's essential files, conducting performance tests. Identifying bottlenecks. Documenting all the analysis done. Discussing solutions. Start implementing the changes. Documenting the process.</p>

<p>PULPino is an open-source microcontroller system, based on a small 32-bit RISC-V core.</p> <p>Importing Arduino Libraries to PULPino architecture should introduce both PULP and RISC-V to the massive Arduino community and would benefit in educational purposes.</p> <p>At the end of this project; all Arduino core libraries would be ported and popular Arduino examples should be tested on PULPino core on FPGA to verify the functionality of the libraries, and all work should be published open source for modification by the open source community.</p>