Find open source projects actively accepting contributors. Search repositories, filter by program milestones, difficulty tags, or tech stack.
Use our Orbit AI Matcher to find out! Get instant matching scores based on your developer skills, preferred frameworks, and contribution experience.
Convert your selected open-source project into a winning GSoC, LFX, or Outreachy application using Proposal Studio.
<p>This project will replace the old xml-based metadata system with a database, which will be assisted by a caching mechanism that will allow for a more responsive user interface. The containerized database will also allow for remote access to job metadata, provided the user has access to the host of the container. The project will enhance the process of large scale computing of batch jobs at CERN and other similar organizations.</p>
<p>Implementation of a web client with additional features for open-source Android smartphone forensic tool that'll help digital forensic investigators throughout the life cycle of digital forensic investigation.</p>
<p>GNU social is a communication software used in federated social networks. This project will: • Improvements on the Federated Requests Queue (which will be shared between OStatus and ActivityPub) • Improvements of the OEmbed plugin • Improvements on the Image Systems • Temporary posts • Removing posts with no engagement • Implementation of a circuit breaker</p>
<p>An open source software project should allow everyone to participate. Inherently this can cause the project to grow over the boundaries of its initial scope rapidly. In Pocket Code this has caused the components to grow into each other, which in turn has made it difficult to maintain a stable and adaptable architecture. I would like to rewrite the existing code into a clearly structured and clearly separated architecture with focus on the distinction between UI representation and functionality.</p>
<p>Details of deep learning models and their performance are, unfortunately, often published without accompanying implementation code. Those models that come from speech recognition seem to be particularly susceptible to this phenomenon. The present project proposes to create a working implementation of a speech recognition model using the Flux library for the Julia programming language and contribute its code to the Flux model zoo. The model to be implemented is Zhang et al.'s (2017) model from their paper "Towards end-to-end speech recognition with deep convolutional neural networks." Due to being implemented using only convolutional layers, this model will be lighter to train than previous models that have used heavier recurrent layers, while still achieving state-of-the-art performance. Having a working implementation of this network will be a step forward in opening the culture of automatic speech recognition. As a result, newcomers to the field will have a recent example to look at for inspiration, which is paramount because there are not many novice-friendly resources available for doing speech recognition research.</p>
<ul> <li>Measure DebugInfo loss statistics.</li> <li>Find the cause.</li> <li>Fix the culprit transformations starting from the most used areas of the compiler.</li> </ul>
<p>The Page Forms extension for MediaWiki is currently used to create and edit pages using an interface made up of forms that simplify the process. The Forms can be used to add and edit template calls in the pages. This project aims to add a special page that provides a spreadsheet-like interface ( rendered using jsGrid ) that will allow administrators of a wiki to edit multiple pages at once.</p>
<p>Data Retriever automates the tasks of finding, downloading, and cleaning up publicly available data, and then stores them in a local database or as .csv files. Simply put, it's a package manager for data. This allows data analysts to spend a majority of their time in analysing rather than in cleaning up or managing data.</p> <p>The Data Retriever is written in Python. It currently has a command line interface (CLI) and can also be used through an associated R package that wraps this CLI. Adding a native Python interface and a Julia package wrapping the CLI would provide access to the tools provided by the Data Retriever in the three major open source languages for data oriented computing.</p>
<p>Currently with WSO2 Enterprise Service Bus recommended way to connect to IBM MQ is using JMS transport. But this has limitations and sometimes gives issues when integrating.Identifying limitations and provide a better solution is part of this project.</p>
The Julia GPU ecosystem allows users to write array programs that run seamlessly across NVIDIA, AMD, and Intel GPUs. This functional portability relies on GPUArrays.jl, which defines the common interface implemented by all vendor packages. However, several foundational operations currently lack shared implementations. Operations such as reverse, findall, accumulate!, and mapreduce either crash on backends lacking a vendor override or silently degrade to CPU execution by copying data back and forth. A direct source audit of GPUArrays.jl (v11.4.1), CUDA.jl, AMDGPU.jl, oneAPI.jl, and Metal.jl confirmed these gaps: reverse and findall are entirely absent from the shared fallback layer, accumulate! exists only as a scalar CPU fallback, and mapreducedim! is declared as an error("Not implemented") stub, forcing every backend to maintain isolated, vendor-specific implementations. This project resolves these portability gaps by establishing a cleaner, more robust architecture: moving critical GPU kernels upstream to AcceleratedKernels.jl (AK.jl) while refining GPUArrays.jl into a lightweight, pure delegation layer. Building upon the newly introduced KernelIntrinsics API, this approach implements highly optimized, vendor-neutral kernels directly within AK.jl. Because Julia’s multiple dispatch inherently prefers the most specific method, routing AnyGPUArray fallbacks to AK.jl means backends with existing optimized vendor methods remain untouched, while backends missing the functionality automatically receive high-performance GPU execution instead of a crash or silent slowdown.
This project addresses the critical challenge of GPU resource management in ML/AI research environments where researchers often face delays due to inefficient resource allocation. I'll develop a comprehensive CPU-based control plane that efficiently orchestrates GPU resources through four key components: an advanced job scheduler with priority management and preemption capabilities, a comprehensive observability platform with real-time dashboards, a streamlined model deployment pipeline with CI/CD integration, and robust infrastructure automation. By implementing technologies like Slurm for scheduling, Prometheus/Grafana for monitoring, Terraform/Ansible for infrastructure automation, and integrating ML-specific tools like MLflow and NVIDIA Triton, the system will deliver significant benefits to researchers: reduced wait times for GPU access, transparent visibility into resource availability, priority access for time-sensitive research, and a seamless deployment pipeline from training to serving. Deliverables include a functioning scheduler with priority queue support, a comprehensive monitoring system with custom dashboards, infrastructure-as-code modules, a model deployment pipeline with CI/CD integration, and an intelligent alerting system. The complete solution will serve as a reference implementation for academic GPU cluster management, providing reusable components and best practices for similar infrastructure requirements in the ML research community.
In particle physics, a cross section is a measure of the likelihood that particles will interact or scatter with one another when they collide. It is a fundamental quantity and its computation can become time intensive. This project aims to apply state-of-the-art machine learning techniques, particularly transformer-based models, for the generation of symbolic representations of the squared amplitudes in HEP. The proposed sequence-to-sequence system aims to improve upon the existing time-intensive processes of calculations of the squared amplitude . This project builds upon the previous work which includes vanilla models, and extends it further to advanced transformer models.
In wireless networking, research utilizing reinforcement learning is on the rise. In response, ns-3 has introduced additional modules like ns3-gym and ns3-ai to support such research. With the increasing traffic and diverse services, 5G NR has been developed and is supported in ns-3 via 5G-LENA. However, there's a need for more reinforcement learning examples based on NR. Additionally, improving the ease of use of multi-agent reinforcement learning (MARL) approaches aims to enhance the usability of NR Sidelink, enabling distributed operations. In this project, I will improve the interface between OpenGymEnv in ns3 and ns3env in the Python ns3 gym module for Multi-Agent Reinforcement Learning (MARL). Currently, ns3gym utilizes the REQ-REP pattern in ZeroMQ as the interface. By incorporating techniques such as client identity and parallel processing of workers in ZMQ, I aim to develop a MARL interface in ns3gym. Additionally, I will enhance the usability of 5G by implementing MARL approaches in 5G examples, with a specific focus on 5G LENA.
This year's goal aims to enhance the TensorMap web application by addressing technical challenges and introducing functionalities to improve usability and efficiency. Key issues, such as dependency management and server-side complexity, will be tackled by Dockerizing the backend for easier deployment. Fixing the neural architecture design workflow will also be the main focus of this year's contribution. New features planned to be introduced into the project include support for image datasets with Convolutional Neural Networks (CNNs), addition of linear regression models for regression tasks, and dataset preprocessing for a better user experience. Finally, documentation on introducing new nodes for future contributors and users along with improved testing coverage for client-side reliability will also be implemented.
Currently, BRL-CAD has implemented a series of NURBS and Brep operations, and in some cases it works well. But there are many cases to be tested, improved or solved like missing surfaces or solids and optimizing topology of Brep. This project works in Test-Driven Development(TDD), aims to test and develop NURBS/Brep part of BRL-CAD, to ensure correctness of geometry and topology in common and special cases.
OpenPIP (Open Source Protein Interaction Platform) is an open-source software package for hosting protein-protein interaction (PPI) data based on the PSI-MI-TAB format. openPIP provides a PPI database and a customisable web interface that enables searching and visualising the interactions as well as downloading the search results in different formats. The whole system is a ready-to-use virtual machine that can be imported to VMware. This project aims to add enhancements and several new features in various sections of openPIP that are used to upload and download data and files, and visualize the complex interaction on various levels with filters and applications.
I am going to use some already proposed NLP methods to implement sub-word based spelling checker, T5 based grammar checker, and merge open source wav2text application into the project to improve the content-wise accuracy of automatic generated subtitle/caption.
<p>The project aims to monitor and track down in detail how the Alexa Top 500 websites behaves to different Tor clients, whether a particular website is blocking the Tor exit node completely, or throwing <code>Captchas</code> or is it disabling certain functionalities <em>(login, posting etc.)</em>. Would also be looking in for similar observations for all tor exit nodes or any peculiar case <em>(region-wise, specific exit node, etc.)</em>. The project also focuses on how often do these webpages return <code>Captchas</code> or even block the pages thereby limiting the users to browse internet freely and anonymously.</p>
<p>New set of tools will be implemented, for users to effectively select the spectral data in the frequency domain, the data selected will then be used for further processing, e.g. Noise reduction or removal.</p> <h3>Basic deliverable</h3> <p>The brush tool will be served as a basic deliverable, whereas the Hann window will likely be chosen with minimal 75% overlapping for both analysis and synthesis part. Regarding circular-shaped data, Bresenham’s algorithm or Midpoint circle algorithm will be used to estimate the circumference points in the frequency domain, the remaining points will be calculated using the Flood fill algorithm.</p> <h3>Optional features: Overtone selection & data reselection</h3> <p>The overtone resonances will now be estimated and selected automatically for the user-selected fundamental frequency. Moreover, the data selected can be refined using the slider, to determine the threshold for spectral energy.</p>
<p>This proposal aims to make histogram calculation a built-in feature of Boost.GIL. The proposal also adds histogram related image processing algorithms to Boost.GIL like histogram equalization, histogram matching, CLAHE, etc. Apart from that, for the purposes of visualizations of the plots, a new extension will be provided for 3D histograms. The existing code in GIL will be updated with the new features implemented.</p> <p>All the code provided will be completely generic such that it works with any container type and any image type supported by GIL. It will also extend support for using Boost.Histogram as a container class. Apart from the histogram related contrast enhancement algorithms implemented, tutorials will also be provided for other contrast enhancement algorithms using GIL features to complete the contrast enhancement suite in GIL. The proposal also aims to add benchmark tools to perform quality tests of the available image processing algorithms.</p>
<p>Enhance scheduling in Ceph’s testing framework, Teuthology, to enable ease of running large scale tests with reliable prioritizing of different tasks in such a manner that fairness and efficiency is achieved with minimum turnaround and maximum throughput. To that effect, replace the existing worker framework by developing a dispatcher responsible for getting jobs from a queue, allocating nodes for these jobs to run on and executing them.</p>
<p>Open science is a movement that promotes open access to research data, publications, and scientific collaboration. It promises to increase transparency, provide reproducibility by reusing scientific datasets and increases trust in the scientific collaboration. Blockchain fits the mindset of open science because of its transparency and decentralized nature.</p> <p>This project aims to create a Hyperledger Composer/Fabric based "Scientific Journal" which could be used to share all the research objects that were produced during the research. These objects (Docs, Videos, Images, Code, Data sets) will be stored in a decentralized manner using IPFS (Inter-Planetory File System).</p>
<p>Improve current functionality of the OwnMailbox personal email server, implement support for a GPG Key Server and HSTS. Furthermore, thoroughly document the whole process, ensure the stability of the code and gain valuable knowledge on linux, cryptography and scripting.</p>