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.
Our "Support Zvk in T1" project aims to upgrade the T1 processor by adding Zvk, a cryptographic feature from the RISC-V architecture, enhancing its data processing and security capabilities. We'll start with researching how Zvk can fit into the T1's design, then move on to detailed planning and design. The heart of the project is the RTL design and implementation, where we'll turn our plans into a functional model. Testing ensures everything works perfectly, leading to the development of a demo application to showcase the new features. Documentation throughout will provide a blueprint for future advancements. This enhancement will not only improve the T1 processor's performance but also its ability to securely handle data, setting a new standard in the field.
An important part of the analysis pipeline of high energy physics experiments is the generation of expected data from first principles. For decades, sophisticated monte carlo simulations have been performed, which are computationally very expensive and therefore pose one of the major bottlenecks in data analyses at CERN experiments. In recent years there has been an effort to explore the application of novel deep learning techniques to speed up the simulation by means of generative models. With the rise of Noisy Intermediate scale Quantum Computing (NISQ), the question comes up, if quantum machine learning can enhance existing methods. In this project we therefore explore the applicability of quantum machine learning to event simulation by means of a quantum Generative Adversarial Network.
<p>TSP is a Ruby on Rails based application to manage the requests and reimbursements from travel help programs of free software organizations. Originally developed for the openSUSE Travel Support Program it was adopted recently by the KDE e.V. Travel Cost Reimbursement initiative.</p> <p>However currently its lacks some features and is less flexible. This project is about making the tool more useful for KDE e.V. and possible future adopters. The new features such as adding an event organizer to manage the events, a feature to send email to the event participants. It also include migrating the UI framework Bootstrap to the latest version and to move the DSL from the models to the configuration files.</p>
This project completes the architectural transition of the HPX parallel runtime to the C++26 Sender/Receiver (S/R) model. Building upon the recent cleanup of legacy execution code, I will advance HPX from an "adapter-based" architecture to its "Native Architectural Phase." This involves refactoring the core execution boundaries of make_future and keep_future to remove internal implementation dependencies (such as the __loop_ pointer) and porting distributed subsystems like MPI and CUDA to be native, first-class stdexec citizens. My primary objective is to eliminate technical debt and achieve zero-overhead integration with the C++26 execution standard while ensuring full backwards compatibility for all existing HPX applications.
Parameterized packages will provide users with a simple way to configure many aspects of packages, à la Gentoo USE flags. This project will add two new values to the Guix package record, "parameters" and "parameter-transforms" which will let package maintainers declare alternative versions of packages that will be built with or without specified features. For example, a user running Guix on a server may choose to build their packages without X11 and Wayland support, and they will now be able to do so easily with the help of these parameters. This project will also implement the logic and user interface for parameters, and specify a number of parameter symbols to serve as examples.
This project aims to address performance limitations of the dirichletprocess R package by reimplementing its computationally intensive core sampling algorithms in C++ while maintaining the existing R interface. The current pure R implementation faces significant scaling challenges with larger datasets. By leveraging Rcpp and RcppArmadillo for seamless integration, I will optimize the MCMC sampling algorithms—particularly the Chinese Restaurant Process sampler and cluster update functions—which are inherently sequential and difficult to vectorize in R. This reimplementation will significantly enhance package performance while preserving the flexible, object-oriented design that distinguishes dirichletprocess from other implementations and allows it to serve as building blocks in custom hierarchical Bayesian nonparametric models.
<p>A social network evolves over time through the creation or deletion of ties among a set of actors.The volatile nature of social ties provides a strong platform to identify the dynamic network structure. This change of structural patterns can be well represented by the existence of dynamic graph objects Further, they can be enriched with the temporal information of social ties to define the recurrent subgraphs of interest. This would yield important insights about the correlation between patterns of ties in a social network. Major contribution of our study is the proposal and implementation of a scalable data structure to handle online and offline dynamic network objects. Hence, it can be used for large scale graph modeling, analysis and visualization purposes.</p>
<p>Apache Gossip is a system which can be used to build a distributed peer-to-peer network using the gossip protocol. Currently Apache Gossip is under its initial development stage where it supports basic data replication with CRDT, fault detection and multiple datacenter support. One of the requirement for this project is to support event based model to notify whenever the per-node data or shared-data get changed. Furthermore, the system is required support distributed locking mechanism in order achieve various tasks. For example dividing a task between nodes or electing a temporary leader. Therefore, the purpose of this project is to design and implement the above mentioned functionalities to Apache Gossip project.</p>
DebNet models the Debian archive as a graph to identify critical packages maintained by too few people. Using data from the Ultimate Debian Database (UDD), it builds a package dependency graph and a maintainer-package graph to compute practical metrics like the Bus Factor, Fragility Score, and Dependency Impact for every source package. The goal is to answer questions like "which packages would cause the most damage if their maintainer became unavailable?" and "where should new contributors focus to help the most?" The deliverable is a FastAPI backend that serves all computed metrics through a REST API with filtering by release, maintainer, and package. This includes bus factor per package, maintainer load distribution, staleness detection, and orphan risk scoring.
<p>Software development is undergoing a gradual increase in pace. The direct result of this is more software vulnerabilities. Countless vulnerabilities are reported and published on different mediums. This data needs to be collected in one place. VulnerableCode aims to fulfil this need.</p> <p>This project enables VulnerableCode to precisely map vulnerabilities from National Vulnerability Database to the vulnerable packages, down to obtaining their package URLs. To do so the current data models are tweaked to capture more details. Data from more ecosystem-specific advisories is collected to minimize reliance on the National Vulnerability Database as well as to get more details. Finally, to allow the community to view and curate the data in a human-friendly way, a basic frontend for VulnerableCode is implemented.</p>
The goal of this project is to help strengthen the Concordia framework by developing and open-sourcing a collection of new language model agent architectures. The project is meant to iterate on the work done by contestants during the Concordia Contest at NeurIPS 2024. While the Concordia Contest 2024 has concluded, the need for well-documented and diverse examples of cooperative agents in the Concordia ecosystem remains important for the ongoing research in cooperative AI. The main benefit being that the new agents created will help lower the barrier of entry for new researchers and engineers, with the hope of inspiring more people to contribute unique approaches to using the framework and speed up the research being done in cooperative AI by providing practical and reusable agent implementations.
LibreOffice's UNO (Universal Network Objects) component model currently lacks support for the Rust programming language, creating a barrier for Rust developers who want to extend LibreOffice or use its functionality. UNO enables cross-language communication and extension capabilities, but is limited to C++, Java, and Python. This project will create a complete Rust language binding for UNO, bridging the gap between Rust's modern safety features and LibreOffice's powerful document processing capabilities. By developing FFI bindings to UNO's C API and providing idiomatic Rust wrappers, this project will enable Rust developers to create LibreOffice extensions and integrate LibreOffice functionality into Rust applications.
The goal of this project (Enext) is to unify and modernize the eventyay-tickets architecture, with a next-generation modular system that integrates ticketing, talk management, and video services under a single, maintainable Django framework. The project focuses on refactoring components, introducing containerized workflows, and improving developer efficiency with consistent configurations and plugin capabilities. Key objectives include: - Unified Event Models and Architecture Refactor - Integration of Tickets, Talks, and Video Components - Plugin Standardization (Badge, Exhibitors, OutputPDF, etc.) - Enhanced CORS, Middleware, and Static File Handling - Docker and Gunicorn Deployment Setup - Performance, Logging, and Celery Worker Integration - Testing, Documentation, and Developer Experience Improvements
<p>Understanding actions and interactions of humans from the RGB-D sensor input can significantly improve cognitive functions of robots and help safely and smoothly incorporate them in the world of humans. Human activity recognition component will thus be a significant addition to the functionality of RoboComp. Human activity recognition in the video is an interesting and challenging task and recent research shows that there are different ways to address the spatial properties of human actions and their temporal dynamics. In the course of this project I plan to start with the simpler LSTM architectures and iteratively test and improve the model to achieve state-of-the art results on the selected datasets with the final goal of providing ready-to-use RoboComp components.</p>
The AI-Powered Media Caption Generator is a smart Drupal module designed to automatically generate engaging and customizable captions for uploaded media using artificial intelligence. It leverages Google Vision API for image understanding and OpenAI along with Hugging Face Transformers for natural language captioning. Users can tailor the captions by choosing the tone, length (0–2000 characters), keywords, and emoji usage. A Python-based backend (using FastAPI or Flask) powers the AI caption generation and offers API support for optional integration with Drupal. The user interface will be built directly within Drupal to ensure a seamless media management and captioning experience.
Text embellishment plays a vital role in enhancing communication, fostering engagement, and making written content more impactful. This project aims to create a Rocket.Chat App that enables users to emojify or embellish their text messages using LLMs (large language models) and prompt engineering. The AI-generated text will be presented through an interactive message interface, with action buttons such as send, redo, edit, and copy, visible only to the user. Depending on the action button clicked, the user can send the message to the channel, regenerate the AI-generated message, edit the message, or copy it, respectively. This app adds a new dimension of creativity to enhancing communication within Rocket.Chat.
<p>Elichika and ch2o are python to ONNX experimental compilers for Chainer ML framework (Elichika would eventually replace ch2o). Given a Chainer model, it parses the python source to get Abstract Syntax Tree and uses it to generate ML framework independent ONNX graphs. Currently neither of the compilers support parsing python jump statements. Moreover, several Chainer functions supported by ch2o are currently unavailable in Elichika. The objective of this project is to extend support for parsing break, continue and pass statements in both Elichika and ch2o. Additionally, it also involves adding Elichika support for Chainer functions and links existing in ch2o but not in Elichika such as sigmoid, max_pooling_2d, BatchNormalization, etc.</p>
<p>This project is aimed towards building a framework which monitors incoming ATLAS Computing Operations data for anomalies and then autonomously, acts on this information by either solving the problem or by proposing the best solution. The solution proposed for this has at its heart two machine learning models-</p> <ul> <li>One a recurrent network to actually predict anomalies in the incoming real-time data. These predicted anomalies will then be analysed to come up with a list of potential solutions to the problem</li> <li>Second a reinforcement learning algorithm to rank the proposed solutions based on a feedback. This algorithm will provide increasingly efficient outputs over time, thus improving the overall efficiency of the framework.</li> </ul>
<p>Effective Quadratures is an open source library used to generate polynomials for parametric computational studies. Among the applications of the library is the ability to perform polynomial regression on a dataset. I propose an implementation of a polynomial regression tree class in Effective Quadratures which will offer greater accuracy when extrapolating compared to the current polynomial regression class.</p> <p>By implementing a polynomial regression tree class I believe that I will help demonstrate some of the benefits of using orthogonal polynomials in the context of model trees. For example users of the library will be able to calculate sensitivity indices and moments over subdomains of functions and datasets. This has the benefit of providing more information about the dataset than was previously possible in the library.</p>
EduAid revolutionizes quiz creation in today's digital education realm with its intuitive interface and advanced AI. Effortlessly generating quizzes for educators and learners, it offers customizable question types and difficulty levels, catering to diverse learning needs. Powered by lightweight Large Language Models (LLMs), EduAid crafts personalized quizzes from input content, ensuring relevance and engagement. Teachers easily assess student understanding, while learners effectively reinforce concepts. In essence, EduAid is a game-changer in educational technology, empowering users to create engaging quizzes effortlessly. Whether shaping assessments or reinforcing knowledge, EduAid delivers precision and ease, enhancing learning experiences significantly.
Problems faced while developing labelled data: 1. Annotations made by a single person can include their biases into the data and eventually into the model 2. There are subjective tasks at times, which might not have a clear answer associated with them 3. The size of the dataset might be huge enough for a few people to annotate To resolve this, a solution is to let multiple people annotate the data and then aggregate the annotations based on a consensus. Crowdsourcing has become an essential paradigm for efficiently labelling large datasets, especially in the era of data-driven ML solutions. However, the challenge arises in aggregating the labels from a diverse pool of annotators, each with varying reliability. This project aims to develop solutions to handle the above problems and integrate their solution into CVAT. It
Investigative work on extending OpenCV. Researched state of the art and emerging 3D multimodal scene representations and use cases. For eg, queryable NeRF methods, semantic embeddings etc. This led to spearheading and designing 2 new initiatives: NerfNet & CognitiveStudio. NerfNet is an open platform for gathering large scale data a la ImageNet for 3D scene representations. It goes end-to-end from images and metadata to 3D scene representations with support for various configs, NeRF models etc. CognitiveStudio is a NerfStudio fork focused on modularity and multimodal integrations Also developed PoCs for LLM agents for converting natural language to NerfStudio API calls and for gathering metadata for NerfNet.
<p>Boost.Real aims to provide a data-type that can represent all types of computable real numbers and provides the flexibility of performing arbitrary precision arithmetic. I propose to optimize the space and time by reducing the dag structure and by various arithmetic optimizations in operation. I propose to optimize space and time in addition and multiplication operations by making adjustments to Karatsuba multiplication and other helper functions. I propose to improve the Champernowne Irrational number and pi's digit computation and add e and golden ratio to irrationals as well. I propose to optimize mathematical functions like exponent, logarithm, trigonometric and add inverse trigonometric functions. After this, Boost.Real may be ready for peer review.</p>
This project will result in functionality for both TARDIS and STARDIS to identify and display a user provided list of atomic and molecular lines to existing spectral visualizations. This will be done by leveraging information stored in the simulation's energy packets to identify source species. This is critical work as observed spectra often have blended line features due to the high ejecta speeds broadening neighboring lines into each other; However, in the simulated spectra, source packets are known and thus blended lines can be disentangled. This work will enable simple comparisons with published, labeled spectra; identification of unlabeled features in newly observed spectra; a physical understanding of new models; and development of publication-ready plots.