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 focuses on refactoring code to kotlin and implementation of MVVM (Model-view-viewModel) architecture for some of the UI components of the application. Implementing some of the recommended Android architectural components along with kotlin refactoring to enhance robustness of existing codebase and make it more testable and maintainable for future development.</p>
<p>Packit-as-a-Service is a GitHub App that helps developers continuously ensure that their project works in Fedora. Packit Service validates pull requests by building that project in Fedora so that others can install and test those built rpms before merging.</p> <p>A very barebones dashboard for packit-service exists. In this project, I want to convert it into a full-featured dashboard/information-center for packit-service. This is a twofold process. We need to properly build the API so that data can be retrieved from packit-service. Then, we need to use that data and build the dashboard itself.</p>
<p>The goal is to implement a reliable pipeline which can take a raw RGB video as input and output information such as whether there is a hand gesture and what is the gesture. The implementation would be applied to a data set of Ref Hen Lab, having around 600,000 hours of television news recordings in multiple languages.</p>
<p>Implement sub-exponential factorization algorithms - (1.) Lenstra's Elliptic Curve factorization Algorithm and (2.) Self-Initializing Quadratic Sieve with Single Large Prime Variation Also implement elliptic curve primality test</p>
<p>Reinforcement learning is a class of machine learning methods that uses the reward signal of the environment to infer optimal actions. Reinforcement learning is used in various fields, including robotic control, conversational bots, resource management, personalized recommendations, and games. Two algorithms are widely used as baselines in reinforcement learning: Deep Q-Networks (DQN; Mnih et al., 2015) and Proximal Policy Optimization (PPO; Schulman et al., 2017). This project will add these two algorithms to Swift for TensorFlow’s open-source repository of models. With the addition of these modules, the users will be able to easily use these algorithms on their problems and rapidly prototype their reinforcement learning algorithm using Swift for TensorFlow.</p>
<p>The goal of the project is to fish the proposed help wanted issues in PiPot GitHub branch, and set up Travis CI for PiPot. Along with feature development and unit testing, we also want to do more research on honeypot, to see whether we can integrate more features/service into the current PiPot.</p>
<p>There is not yet any way in R to fully leverage the power of stochastic gradient algorithms for fitting generalized linear models (GLM). The overarching goal of this project is to develop the R package sgdnet into a mature state for the implementation of the efficient SAGA algorithm for elastic net-penalized GLMs, targeting the big data setting where observations greatly outnumber variables. It will result in a CRAN submission.</p>
<p>Livechat is a package that adds the ability to embed a pop-up support chat to your website. For example when you want to interact with users visiting your website and help them with their problems, then this feature of RocketChat helps you with that. All you need to do embed a script on to your static website. Live chat agents who wield this tool use it to make customers happy and satisfied.</p> <p>This project adds real-time monitoring to livechat which will be an important feature as this will be helpful for livechat agents in the following way:-</p> <ul> <li><strong>More information about the user</strong>:- Track important information related to visitors, such as the <em>current page, location, time on site, local time</em>, etc, even though the visitor has not yet started a conversation with an agent.</li> <li><strong>Edit Info Panel for livechat with the user</strong>.</li> <li><strong>Increase Response TIme</strong>:- Agent will be able to start a conversation with a specific visitor directly from the list, improving the proactivity of the Live chat team.</li> <li><strong>New Admin Panel providing more info of users</strong></li> <li><strong>Add sentiment analysis on user messages</strong>:- When chat will be closed, sentiment analysis will be applied on user messages and shown on the admin dashboard for future use.</li> </ul>
<p>TrackPal is a public based mobile application that will be implemented using react-native. And the components that are in operation with this application, will be created based on the Go-social components. By using this mobile application, users can share their locations when they are on a bus or train and that shared location will be updated when it moves and other users can see the shared details. This will solve the problem with users have, so that they can find the bus's or train's current location by just tapping on a route or train number on the screen.</p>
<p>A Slack integrated Github automation bot for handling and checking Github workflow.</p>
<h3>CCAligner</h3> <p><em><strong>Word by Word Subtitle Synchronization Tool</strong></em></p> <p>The usual subtitle files (such as SubRips) have line by line synchronization in them i.e. the subtitles containing the dialogue appear when the person starts talking and disappears when the dialogue finishes. This continues for the whole video. For example :</p> <pre><code>1274 01:55:48,484 --> 01:55:50,860 The Force is strong with this one </code></pre> <p>In the above example, the dialogue #1274 - <code>The Force is strong with this one</code> appears at <code>1:55:48</code> remains in the screen for two seconds and disappears at <code>1:55:50</code>.</p> <p>The aim of the project is to tag the word as it is spoken, similar to that in karaoke systems.</p> <p>E.g.</p> <pre><code>The [6948484:6948500] Force [6948501:6948633] is [6948634:6948710] strong [6948711:6949999] with [6949100:6949313] </code></pre> <p>In the above example each word from subtitle is tagged with beginning and ending timestamps based on audio.</p> <p>Project for <em>CCExtractor Development, GSoC 2017</em> by <em>Saurabh Shrivastava</em>.<br> <em>saurabh.shrivastava54@gmail.com</em><br> <em>saurabhshri.github.io</em></p>
<p>This project is implementation of disk controller emulator for bhyve. I'l implement is NVMe controller. NVMe is defined by Non-Volatile Memory Host Controller Interface Working Group. NVMe has some advantage compared with another host interface, for example, scalable, small command, large command buffer and so on. It can say with disk controller emulator for virtual machine too. Small command can reduce MMIO emulation, thus, we can obtain performance improvement. Bhyve has mechanism of pci emulation. I'll use this one for implementation of NVMe controller emulator. This is similar to pci ahci. The pci ahci controller emulator is already implemented in bhyve, therefore, I can reference this for my project. I divide task for finishing project. First step, I'l implement basically function, for example, minimum queues and commands. When I finish implementation of fundamental function, I'l take more advanced function. For example, namespace queue, shared queue, fused operation and so on.</p>
<p>WikiPathways is an open and collaborative platform for sharing biological pathway information. It features an interactive diagram that aids the user in understanding the pathway. However, the educational capabilities of these diagrams can be improved. Studies show that when information is given in steps, with dynamic visual aids, audience attention and understanding increases dramatically. Ideas presented graphically, as a “story", are far easier to comprehend and remember than static information. Thus, step-by-step presentations of pathway diagrams would dramatically enhance the educational capabilities of WikiPathways.</p> <p>In this proposal, a new "Pathway Presenter" is outlined that will enable biologists to create dynamic presentations of the pathways in WIkiPathways. This will allow for the creation of succinct and easily comprehensible diagrams that can be shared and embedded across multiple platforms.</p>
<p>Frontera is a distributed web crawling framework, it acts as a crawl frontier, dealing with the crawling strategy to be followed. The goal of this project is to port frontera to python 2/python 3 dual compatibility, and improving test coverage and other improvements along the way.</p>
<p>In Data Science acquiring insight from data is what matters and any good data analysis tool must have the key functionalities required to extract information from data. Now, it’s needless to say how ubiquitous categorical data is in Data Analysis but they are not currently supported in Daru and operations on them like regression is not supported in Statsample and Statsample-glm. This shortcoming of Daru prevents data to be easily and to its full extent analyzed. This project aims to achieve just that.</p> <p>This project has broadly two goals:</p> <ul> <li>First is to efficiently store and manage categorical data in a Data Frame. This will be done with the help of a new data type CategoricalData and a class CategoricalIndex dedicated for this purpose. It also involves easy visualization of categorical data with help of plotting functionality.</li> <li>Next goal is to update various statistical analysis such as regression in Statsample and Statsample-glm taking categorical data into account.</li> </ul> <p>On accomplishing both of these tasks, one would be able to see Data more clearly.</p>
This project aims to extend the existing Marine AR Module Builder into a more accessible, scalable, and educator-friendly toolkit for creating augmented reality learning experiences in Unity. The primary objective is to reduce the technical barrier for educators and developers by introducing intuitive editor tools, structured workflows, and reusable systems for building marine AR modules. The proposed system will include drag-and-drop environment upgradation, streamlined actor placement and configuration, and reusable behavior templates for movement, interaction, and ecological logic. To improve flexibility and maintainability, modules will support extensive JSON/script-based definitions, enabling structured and streamlined content creation. Validation tools will be implemented to detect missing assets, invalid configurations, and performance issues, ensuring reliability and usability.
This project will improve the RenAIssance OCR pipeline for 17th-century Spanish historical documents by strengthening preprocessing, removing marginalia, optimizing TrOCR inference, and adding a flexible correction and deployment stack. The plan is to convert raw scanned pages into cleaner, layout-aware inputs, improve printed-text OCR, extend the system to handwritten OCR with the available handwritten datasets, and integrate local LLM/VLM-based correction with external fallback support when needed. The final deliverables will be a modular OCR pipeline, a deployed Flask application, benchmarking against existing OCR tools, and a paper-style technical report with reproducible documentation.
kornia-rs currently supports only CPU-based computation for image processing operations like resize, warp, and color transforms. This project implements a GPU backend for kornia-tensor and kornia-imgproc, motivated by a concrete Bubbaloop use case: a real-time camera pipeline that needs faster preprocessing to make VLM inference feasible on edge hardware. The work includes a feature-gated GpuAllocator with explicit host-device transfer semantics, allocator-based backend dispatch for tensor operations, and GPU kernels for resize, warp_affine, warp_perspective, and RGB <-> Grayscale conversion using CubeCL.
This project aims to enhance DeepForest by integrating and extending the active learning capabilities from Weecology/BOEM into its training pipeline. The goal is to enhance the efficiency and scalability of tree and bird detection in aerial imagery, enabling smarter sample selection and reduced annotation effort. By using advanced sampling strategies such as uncertainty-based and diversity-based selection, users will be able to automatically generate predictions, identify the most valuable samples for annotation, and streamline the entire labeling process. The project will also focus on building an intuitive workflow, writing clear documentation, ensuring strong test coverage, and creating user-friendly interfaces. Ultimately, this work will help reduce manual annotation efforts, speed up training, and improve model performance, making DeepForest more adaptable.
This project addresses the challenge of making advanced Transformer-based OCR technology for historical documents accessible to non-technical users on standard hardware. Despite significant advances in OCR accuracy for complex historical texts, current solutions remain inaccessible to many historians, archivists, and researchers due to technical barriers and hardware requirements. The solution involves developing an optimized, user-friendly OCR system specifically for historical documents through three main approaches: (1) implementing advanced model optimization techniques for efficient CPU inference, including kernel fusion, weight pruning, and knowledge distillation; (2) creating cross-platform deployment options for mobile devices and web browsers; and (3) developing an intuitive application interface with preprocessing controls and interactive correction capabilities. Deliverables will include: 1) A quantized OCR model achieving at least 4x speedup on CPU with <1% accuracy loss. 2) Browser-based inference capability: An enhanced Streamlit application with batch processing, export options and 1 click OCR. 3) A benchmark suite comparing performance across hardware configurations 4) Comprehensive documentation and tutorials for non-technical users This project will democratize access to cutting-edge OCR technology, enabling broader participation in historical document digitization and analysis without specialized technical expertise.
In the current testing workflow, after pushing a branch to ceph-ci the developers can run integration tests using teuthology only after Ceph binaries are built, which takes a few hours. With this project, they will be able to queue commands using teuthology-api before the Jenkins build is completed. Once the build is complete, the queued commands will get executed automatically. New endpoints for auto-scheduling and saving commands commonly used by a developer will be added to teuthology-api.
Idea based on https://wiki.openstreetmap.org/wiki/Google_Summer_of_Code/2024/Project_ideas Ferrostar is the new cross-platform turn-by-turn navigation SDK. Currently, it’s only targeting iOS and Android platforms. By extending Ferrostar's functionality to Progressive Web Applications (PWAs), we aim to unlock its potential in diverse scenarios. PWAs offer a lightweight, accessible platform for delivering navigation services without the need for specialized native development teams. Additionally, PWAs can greatly broaden the versatility of navigation, such as expanding accessibility to turn-by-turn navigation on handheld PCs and non-Android embedded devices, and some scenarios like Apple App Clips (basically Apple’s rebranded PWAs) which are convenient when downloading apps being clunky (for example, navigating in a park).
This proposal aims to implement VirtIO sound driver for Haiku. Haiku’s performance in virtualized environments will be significantly improved, instead of relying on emulated sound cards or other drivers.
This project intends to allow Picard to post process cover art images obtained from the providers. There should be provider independent options to restrict image sizes, resize images if it exceeds a maximum allowed size or automatic format conversion. Also, there should be image post processing plugin hooks to allow plugins to do additional image manipulations.