Explore Projects

<p>The objective of this project is to minimize mlpack library algorithms to fit into resource constrained devices such as Raspberry PI, Arduino, or VoCore</p>

<p>The rendering component of Gazebo has been moved to its own library, Ignition Rendering, and is now part of the <a href="https://bitbucket.org/ignitionrobotics/ign-rendering" target="_blank">Ignition Robotics</a> project. The works involves porting a list of sensor visualization features from Gazebo to <a href="https://ignitionrobotics.org/" target="_blank">Ignition Rendering</a>. These include laser, sonar, wrench, contact visuals and more. The student will create these visuals in Ignition Rendering as loadable plugins, implement any features needed to create these visualizations, and demonstrate the results using simulated sensor data generated by Ignition Sensors.</p>

<p>Hello! I am Rucha Deodhar. The task I've picked is to add RETURNING clause to INSERT statement which will return the listed fields after the insert query runs successfully. This will make it library first and also might improve compatibility with Oracle RDBMS. It also eliminates the need of additional query. This feature already exists in other DBMS like PostgreSQL. More information about my project can be found in the proposal. Thanks!</p>

<p>This game will encourage female teenagers to develop their own games in the Catrobat app. As a tutorial game it is the gateway to Catrobat and its appeal and its learning effects will decide, for many users, weather or not they become active users of the app. For many teenage girls, this will be their first experience with coding and, if it is designed successfully, it may direct their interests towards programming, a field they may have never considered otherwise. The effectiveness of the game will be evaluated in the context of a study directed by the Institute of developmental psychology of the Karl Franzens University, Graz.</p>

<p>Iodide is a browser-based notebook which works with absolute zero configuration. Iodide notebooks are designed for scientists so that they can work flexibly and share their work without any hassle. Iodide is still in its early alpha stage. This project aims to extend the functionality of the notebook; making the notebook code more robust and performant; helping to shape the user experience; and creating example notebooks.</p>

<p>PSLab (Pocket Science Lab) is an Open Source hardware device (open on all layers). The Android App for PSLab aims to run PSLab device in android smartphones.This proposal offers to upgrade PSLab Android App by implementing all the leftover functionalities in the Android App.</p>

<p>Distributed and secure communication for the sake of user's privacy has never been as much needed as of now. That's why I feel a great pleasure to work on <a href="http://opendht.net" target="_blank">OpenDHT</a>, a critical library used by <a href="http://ring.cx" target="_blank">Ring</a>, a project soon to be imported in the debian software. This summer, thanks to the Google Summer Of Code program, I will work on developping new functionalities in OpenDHT aiming at reducing overall generated traffic, maintaining and optimizing critical parts of OpenDHT, optimizing data persistence solution on the network and use TCP protocol instead of UDP for OpenDHT.</p>

<p>The project deals with downloading of subtitles from popular online TV Services like Netflix, BBC, Amazon, Hulu. The project aims to perform this task without the need for the user to interact with the browser.</p>

My name is Edward. I'm a Master’s student at the University of Colorado Boulder in the US and I am passionate about Apache Fineract’s mission to improve the availability of financial services to people all over the world. Apache Fineract provides core banking APIs, but lacks a community-maintained layer for end-users. The former Self-Service API was removed, leaving vendors to build their own frontend from scratch. To support this new frontend, I am proposing a new Backend for Frontend (BFF) and associated frontend that enables improvements to end-user security, reliability, performance, and auditing. The focus of this project is to build a POC that is well-tested and documented so we can build a "release-ready" solution in the future. Deliverables include: 1) Standalone Spring Boot BFF application with its own user model and JWT-based authentication 2) Feign client integration with Fineract's core API via a dedicated service account 3) API Aggregations for User data summaries, account information, loan and savings views, and transaction capabilities 4) User registration flow with OTP and Identity Verification 5) User login flow with 2FA 6) Auditing for Basic Frontend Concerns 7) Open Banking Authentication Flow 8) A working frontend to demo these features Some stretch goals include but are not limited to: 1) MPIN for transaction authorization 2) Rate Limiting 3) HSM key storage

The eventyay presale and ticketing flow is currently built with Django templates and jQuery, resulting in full page reloads between every step of the ticket purchase journey. This project will migrate that flow into a Vue 3 Single Page Application using Vue Router for client-side navigation, centralized state management for cart and checkout state, and a unified API layer interfacing with the existing Django REST Framework endpoints. The migration follows an incremental approach where Django templates remain active throughout, keeping each phase independently reversible. Deliverables include a fully functional SPA covering event discovery, product selection, cart management, voucher handling, and checkout, along with route guards, performance optimization through lazy loading, frontend tests, and contributor documentation.

This project aims to develop a pipeline for producing calorimeter shower datasets in point-cloud form for a given detector and various types of particles. Its central goal is not only to compress detailed Geant4 shower information into a compact point-cloud representation, but to do so while preserving the physics content needed for the downstream. The work will proceed in two major stages. First, electromagnetic (EM) showers will be used to establish the pipeline, optimize point-cloud construction, and define a robust validation protocol. Second, the same framework will be extended to hadronic showers, which are more challenging due to larger fluctuations, invisible energy, delayed components, and a stronger dependence on material and interaction modeling.

Eclipse Theia is used by developers to build unique desktop and cloud IDEs. The framework's developer experience is becoming more and more important for both external users and internal contributors as its use grows. The goal of this project is to enhance the framework in four key areas: OpenTelemetry Evaluation, Detachable Views & Window Management, Startup Performance Profiling, and Unified Logging & Observability. The project has a variety of goals. These include addressing CSS propagation issues within detached Lumino widgets, migrating core modules to the ILogger standard, reducing Largest Contentful Paint (LCP) during application startup, and creating an architectural proof of concept for distributed tracing that covers frontend and backend RPC boundaries.

The Unikraft ports for popular applications like Redis and Nginx are currently outdated and fail to compile with the latest kernel core. This project solves this problem by updating the application glue code and implementing missing system calls in the ukposix compatibility layer. The main deliverables are fully working, stable, and benchmarked unikernel versions of Redis and Nginx. This work will also improve the overall POSIX support in Unikraft, making it easier to port other applications in the future.

This project aims to enable testing of Guarded Control Stack (GCS) functionality within CRIU (Checkpoint/Restore In Userspace) on AArch64 by integrating the Armv9.4-A GCS feature into CRIU’s ZDTM test suite through a dedicated test case that confirms whether TPIDR2_EL0 is preserved and still readable in user space after restore. The project leverages recent upstream additions to glibc and Linux kernel that enable GCS support in user space, using Yocto and Arm's Fixed Virtual Platform (FVP) to emulate and test the architecture-specific functionality. This is a first step in exploring how CRIU might need to change to handle GCS systems, starting with test coverage Project Deliverables June 2 (coding begins) - GCS-enabled Environment Verified. Yocto + FVP setup is confirmed working. Compiling and running GCS-marked binaries with glibc tunables active. July 12-16 (Midterm Evaluation) - First GCS ZDTM test added. A basic test that reads verifies TPIDR2_EL0, and explores stack/register consistency across the restore boundary. Integration with --nocr working. Investigation of parasite.c and restorer.c to identify the point at which TLS or control state (e.g., TPIDR2_EL0) could be saved or restored. August 16-August 23 (Final Evaluation) - Full round-trip test of checkpointing and restoring a GCS-using process. All tunables, toolchain flags, and kernel glibc dependencies documented. August 31 (Final Results Announced) - Final cleanup and documentation. Tutorial-style instructions provided on how to replicate setup, test flow, and next steps. Post GSOC - I will remain available to support further maintenance, and help any future contributors pick up.

This proposal aims to build upon and expand the progress made in recent years on Quantum Vision Transformers (QViTs) for High Energy Physics (HEP) analysis. Specifically, the project will adapt and extend the state-of-the-art Particle Transformer architecture, which has already demonstrated excellent performance in jet classification tasks, by integrating it with quantum components.

Optical Character Recognition (OCR) technology has revolutionized document digitization. However, existing OCR tools like Adobe Acrobat and Google’s Pytesseract, they struggle with historical documents due to variations in font styles, image quality, degradation, and layouts. Text recognition from centuries-old works poses a significant challenge due to the complexity of early non-standard print forms, handwritten text and manuscripts. Over the past year, transformer-based models have been fine-tuned to improve transcription accuracy, particularly for degraded and complex historical texts. Training on a diverse dataset that combines expert transcriptions and synthetic data has enabled better generalization across various typographical styles. Building on this progress, we aim to expand the dataset, to help the model finetune to handle handwritten documents as well. Fine-tuning on larger datasets incorporating diverse typographical styles both printed and handwritten and also implementing Language Modeling & Contextual Understanding for post-processing, allowing for contextual corrections based on 17th-century grammar to further enhancing the OCR accuracy. Additionally, we plan to develop and deploy a web or mobile-based annotation tool for historians, researchers, and institutions to validate and refine OCR outputs

In high-energy physics (HEP), automating the symbolic computation of squared amplitudes is essential for predicting cross-sections and validating theoretical models against experimental data. Current transformer-based approaches excel in sequence modeling but falter in capturing the relational structures and physical consistency required for HEP expressions, such as those involving gamma matrices and Mandelstam variables. To tackle this, I propose the Physics-Informed Performer (PIP), a novel framework that integrates the Performer transformer’s efficient kernel-based attention with a physics-informed tokenization scheme and a SymPy-based validation layer. This approach leverages Performer’s ability to handle long sequences, introduces tokens like <SQUARE> and <S> to embed HEP-specific operations and kinematics, and ensures symbolic accuracy through empirical post-processing. By focusing on data representation, PIP offers a computationally efficient alternative to traditional symbolic methods, tailored to the Task 1.2 dataset of amplitude-to-squared-amplitude mappings.

This proposal outlines a detailed, structured, and actionable two-phase strategy for developing an OpenVINO AI PC Model Training Kit. The core objective of this initiative is to enable streamlined, efficient, and scalable AI model training capabilities on widely accessible hardware by fully leveraging OpenVINO’s optimization capabilities and integrating it seamlessly with leading machine learning frameworks such as PyTorch, TensorFlow, and Scikit-learn.

Kubernetes has recently moved its forensic container checkpointing feature into beta. This feature allows checkpointing containers within Kubernetes pods, and, with some additional steps, restoring them on a different node. This approach preserves active TCP connections of the container. However, workloads (e.g., web servers) with established TCP connections to the migrated container cannot make use of this feature, as the container IP changes after migrating to another node. This project implements a P4-based load balancer on the data plane that dynamically reroutes packets to the correct IP address after container migration. The goal is to make the migration invisible to the client, causing only a period of service unavailability, after which communication can be resumed without re-establishing the TCP connection. Deliverables of this project include a P4-based load balancer, control plane code for P4 runtime interaction, a minimal CNI plugin for Kubernetes integration, and performance benchmarks to demonstrate the system's effectiveness.

The project incorporates addition of a new video filter to FFmpeg, PU21 transform. PU21 has been developed to convert absolute high dynamic range (HDR) linear color values into approximately perceptually uniform (PU) values, which can be used with standard quality metrics. This transformation would eventually help evaluate HDR videos based on quality tests known for SDR ones.

The proposed project aims to address the sustainability challenges faced by open-source software projects by using an agent-based modelling and simulation approach. Open-source projects often face issues such as limited resources, difficulty in attracting and retaining contributors, and communication breakdowns, which can hinder their growth and sustainability. To address these challenges, the project will simulate various scenarios and identify the factors that contribute to the success or failure of open-source projects. The project will provide a framework that enables simulating different scenarios to assist project maintainers in making informed decisions that promote sustainability. This approach can offer a more detailed and context-specific understanding of the challenges open-source projects face and develop effective strategies for maintaining their sustainability.

In short, this proposal aims to export more pola-rs bindings to R beyond existing ones offered in r-polars, and make such bindings as efficient as possible. Pola-rs is a highly efficient data manipulation library written in Rust, and it can significantly improve the performance of data pipelines. It would be especially helpful for computationally demanding analysis if such library is accessible in R.

<p>Multi-stage decision-making problems widely appear and pose unique challenges in various real-world applications, e.g., robot control, game play, and healthcare. Consider the clinical trial design as an example. Optimizing the clinical trial design is essential to decrease the cost of drug development, as the average costs are up to 52.9 million US dollars for a Phase 3 trial [1]. The current approach formulates multi-stage decision-making as a general Linear Programming (LP) problem, which however, is computationally expensive especially in large-scale settings. In this project, we aim to develop new customized algorithms and an R package for multi-stage decision-making problems with three key features: 1) It provides a highly efficient solver to tackle a large and important class of sparse LP problems; 2) It provides a solution for multi-stage decision-making problems with Bayes risk constraints; 3) It provides additional functions such as visualization of the optimal decision maps.</p>