Explore Projects

<p>Origami aims to provide Artificial Intelligence as a Service. Presently, deep learning is a very interesting field. However, the issue arises when someone has to setup someone else's code on their machine. Origami aims at solving that problem by allowing researchers to quickly create a demo of their implementation for the world to see. This serves a twofold purpose. It lets other users see if a particular implementation solves their purpose while also serving as a 'portfolio' for the researchers.</p> <p>By the end of the summer, users will be able to search for various demos published by various researchers and thus be able to find an implementation that suits their needs the best. Additionally, Origami will also support REST APIs. This will allow Origami to be used to provide some infrastructure for chat bots which can then be integrated with Slack, Facebook, etc. Origami will also be integrated with EvalAI to allow users to compare their implementation with that of their competitors. Finally, Origami will support other features such as allowing users to upload to/from Dropbox, releasing anonymous demos, a better CMS and predefined templates to create apps to name a few.</p>

<p>This is a user interface to draw and configure deep neural networks and supports import / export of model configuration file from / to caffe &amp; tensorflow deep learning frameworks</p>

<p>The proposal is towards improving the layer support and adding real time collaboration for Fabrik. I propose to add support for some essential missing layers, creating a Dummy layer which will automatically load any unsupported layers, and adding real time collaboration capabilities to the framework.</p>

<p>Linux System Roles is a project related to Ansible, a tool for automating configuration management, application deployment and software provisioning. The goal of linux system roles is to provide a consistent user interface, abstracting from any particular implementation of the linux subsystems, but trying to get the most out of the particular libraries on each one of them.</p> <p>The Network Linux System Role, the one that aims to be improved in this proposal, currently provides a unique configuration interface for network-scripts and NetworkManager. This tool requires many improvements in several areas as it is a relatively new tool.</p> <p>In the proposal, all the issues are addressed, ordered from the tasks that the candidate considered to be better for the start in the project, to the ones that seems more challenging. The main goal of the summer will be to add Pytest support to the integration testing of the project.</p>

<p>Allow any user to keep the changes he does when he's booted into the live system across reboots.</p>

<p>Fabrik is an online collaborative platform to build, visualize and train deep learning models via a simple drag-and-drop interface. It allows researchers to collaboratively develop and debug models using a web GUI that supports importing, editing and exporting networks written in widely popular frameworks like Caffe, Keras, and TensorFlow.</p>

<p>Fedora has an android app which lets a user browse Fedora Magazine, Fedora Ask, Fedocal etc within it. This app is build using the Ionic Framework, Angular and Cordova. Essentially it is a cross-platform hybrid app.</p> <p>In the current form, most of the functions rely on an in-app browser to render content. This project aims to improve the existing Fedora App for Android for speed, utility, and responsiveness, introduce a deeper native integration and make the app more personal for the user.</p>

This project aims to enhance and optimize the existing platform by implementing several important features and improvements. These changes include the addition of domains, tags, and a prize and sponsor page, as well as the introduction of a deregister process, a revamped profile page with a new tabular UI design, a participant team approval feature, and the inclusion of GMT time stamps and many more features.

<p>The rise of reinforcement learning based problems or any problem which requires that an agent must interact with an environment introduces additional challenges for benchmarking. In contrast to the supervised learning setting where performance is measured by evaluating on a static test set, it is less straightforward to measure generalisation performance of these agents in the context of the interactions with the environment. Evaluating these agents involves running the associated code on a collection of unseen environments that constitutes a hidden test set for such a scenario. The goal of this project is to set up a robust pipeline for uploading prediction code in the form of Docker containers that will be evaluated on remote machines and the results will be displayed on the leaderboard.</p>

<p>A tool for the Managers and the contributors to proposal/edit/approve changes as per fedoras change process.</p>

The AlphaOne Labs platform (alphaonelabs/website) is a Django monolith. This proposal migrates it to a serverless Cloudflare Worker architecture (alphaonelabs/learn), decomposing the monolith into a Python CF Worker, D1 SQLite database, R2 object storage, KV cache, CF Pages frontend, CF Queues for async jobs, and Durable Objects for real-time virtual classrooms. As a secondary deliverable, it implements an AI powered Personalised Learning Lab with a Learner Knowledge Graph, spaced repetition (SM-2), RAG pipeline using Cloudflare Vectorize and Workers AI, and three adaptive tutor modes. The migration is not greenfield, the learn repo is already operational. Deliverables: fully migrated platform on Cloudflare, virtual classroom rebuilt on Durable Objects, and a working adaptive AI tutor module.

The DU Event Board currently uses manual inputs and basic validation to manage event data. As the number of events grows, this can make it harder to maintain consistency, avoid duplicate entries, and optimize API usage. Event discovery is also primarily based on simple filters and does not yet include offline access. This project improves how event data is processed and maintained. On the backend, it extends the existing GitHub Actions workflow to automatically fetch event metadata (image and description), detect duplicates using title and date similarity, and perform geocoding only once per event by writing coordinates back to the source file. These changes prevent duplicate entries, reduce repeated API calls, and remove the need for manual data fixes. On the frontend, the project adds full-text search, map-based filtering with a time slider, relative date display, and offline support. It also introduces a guided event submission flow that creates structured GitHub Issues, making it easier for contributors to add events. Together, these changes reduce manual review work and make it easier for users to search, filter, and access events.

Swarm external secrets project enables Docker Swarm services to consume secrets from external secret managers, but it still lacks several features needed for secure and production ready deployments. At present, some providers rely on older credential models, transport security can be improved, and support for newer secret backends and modern authentication workflows is limited. This project aims to close those gaps by making the plugin more secure, extensible, and operationally reliable. The proposed work will focus on adding modern identity-based authentication methods such as JWT and OIDC flows for Vault and related providers, improving TLS and mTLS support with better CA bundle handling, and strengthening secret lifecycle support such as version awareness and rotation compatibility. In parallel, the project will expand provider support by integrating additional backends such as 1Password, Doppler, and Infisical. The deliverables will include secure authentication improvements, stronger transport security, new provider implementations, improved tests and documentation, and better observability so that swarm-external-secrets becomes a more enterprise-ready secret management solution for Docker Swarm.

This project will enable Sugar users to manage Docker Swarm clusters directly through Sugar’s command-line interface (CLI). These enhancements will streamline container orchestration workflows and make Sugar a more versatile tool for developers transitioning from development to production environments. New commands will enable users to perform tasks such as initializing, joining, creating, scaling, updating, and inspecting Swarm services with simplified Docker Swarm Node and Secrets Management.

<p>The MINIX3 has already supported ARM-based platforms such as BeagleBoard. The goal of this project is port the MINIX3 to the RaspberryPi. The task requires changing booting process, because it doesn't allow have more than one platform. Also, some of RaspberryPi drivers such as USB, SDCard, SPI, I2C, etc. must be implemented.</p>

<p>In order to build Redox on Redox, Rustc has to be ported, along with Cargo, and any dependencies (such as LLVM). Additionally, the build process currently requires make and a Bourne-compatible shell, which will have to be ported, or a way to build without them provided.</p> <p>LLVM seems to assume it is running under Windows or Unix. So the code in lib/Support/Unix and lib/Support/Windows likely needs to be reimplemented for Redox.</p>

Improve Miniflare service-to-service RPC bindings and local D1 database schema migration replays.

This project creates a comprehensive pipeline for extracting text and complex tables from scientific literature PDFs. By combining PyMuPDF, specialized table detection models, and Vision-Language Models (VLMs), the system will significantly improve accuracy over traditional OCR methods while maintaining cost efficiency. The pipeline allows users to upload multiple research papers, select their preferred extraction method, and receive structured markdown outputs with preserved document hierarchy. A side-by-side interface will enable human verification and correction of extraction results. The solution will be optimized for lightweight deployment on platforms like Hugging Face Spaces, making it accessible to researchers worldwide while focusing primarily on complex table formats common in life science research.

This proposal integrates IRx with the Apache Arrow C Data Interface (CDI) to enable zero-copy interoperability with systems such as PyArrow and DuckDB. Currently, IRx lacks a complete Arrow export path, supports only int32, and contains a critical issue where finish() hardcodes type metadata, leading to silent data corruption for other types. The project introduces a full Arrow lowering pipeline covering all fixed-width types and UTF-8 strings, implements a correct ArrowArray export mechanism, and resolves type handling through a generalized type_id-based runtime design. This ensures that future type additions remain localized and scalable without modifying shared logic. The outcome is a production-ready integration that allows IRx programs to exchange data directly with Arrow-compatible systems without copying or serialization, enabling its use in real-world data processing pipelines.

HiperHealth's clinical pipeline currently lacks visual diagnostic capabilities. This project adds a modular MedVision skill channel, an external, installable module that brings medical image analysis into the existing pipeline with zero core modifications. The project delivers shared infrastructure (MONAI preprocessing, Grad-CAM explainability, structured Pydantic output models, confidence calibration) and two body-region skills: SkinAnalysisSkill (DINOv2 backbone, ISIC 2018 dataset) and a second body-region skill (Eye/Nail) to validate extensibility. Each skill automatically enriches the DiagnosticsSkill's differential diagnosis via the prompt_fragments mechanism, producing calibrated predictions with saliency maps for clinical transparency. The project includes Fitzpatrick fairness auditing and ≥90% test coverage.

The objective is to enhance the Beagle-Tester framework for BeagleBoard devices by adding mikroBUS support and updating it for mainline kernel testing. This project aims to develop an automated regression test suite for Linux kernel on BeagleBoard hardware, enabling continuous validation within the OpenBeagle CI server. The approach includes developing comprehensive test cases for interfaces such as PWM, ADC, UART, I2C, SPI, and GPIO, integrating automated power cycling capabilities for test farms, building a web interface for real-time monitoring of test results, and optimizing the framework for seamless Buildroot deployment. Key deliverables include an enhanced Beagle-Tester with mikroBUS support, an automated regression testing framework for kernel updates, a GitLab CI pipeline for weekly testing, a modular test farm design with power cycling capabilities, a web-based monitoring interface, and comprehensive documentation with tutorials.

The Greybus module for Zephyr currently exists as an out-of-tree component, making maintenance, testing, and contributions difficult. This project aims to upstream Greybus as an official Zephyr module, improving maintainability, accessibility, and seamless integration within the Zephyr RTOS ecosystem. By upstreaming Greybus, BeagleConnect and other Greybus-enabled platforms will gain native support in Zephyr, removing the need for custom patches or kernel modifications. Originally developed by Google as part of Project Ara, Greybus is a lightweight, high-speed protocol designed for efficient, low-power communication across diverse hardware interfaces. While initially intended for modular smartphones, Greybus has since expanded into IoT and industrial applications, particularly in wireless sensor networks. This project will focus on cleaning up and refactoring the existing Greybus module, revamping its testing infrastructure, and ensuring compatibility with Zephyr’s CI/CD workflows. The successful completion of this work will enable seamless hardware-software interaction, making Greybus more accessible for embedded systems development within Zephyr.

Current SoCs frequently feature heterogeneous remote processor units in asymmetric multiprocessing (AMP) setups, which may be running several Linux or other real-time OS instances. One such SoC is TDA4VM, with heterogeneous multicore support it can run multiple operating systems simultaneously. TDA4VM has Dual 64-bit Arm® Cortex®-A72 microprocessor subsystem at up to 2.0 GHz, Six Arm® Cortex®-R5F MCUs at up to 1.0 GHz, Two C66x floating point DSP, up to 1.35 GHz, 40 GFLOPS, 160 GOPS and 3D GPU PowerVR® Rogue 8XE GE8430, up to 750 MHz, 96 GFLOPS, 6 Gpix/sec. Zephyr is a small real-time operating system for connected, resource-constrained and embedded devices supporting multiple architectures. Zephyr allows for easy handling of multiple configuration options, APIs and external components, and is well suited to structured application development. Another benefit of Zephyr is that it targets some very serious protocol and standard implementations, being e.g. the first open source RTOS to introduce TSN support – by way of Antmicro’s contribution. The rising popularity of TSN in automotive and aerospace applications, and just about everywhere else, could be a very important reason to start using Zephyr in your TSN-capable product. The goal of the project is to add Zephyr RTOS support to run on Cortex R5 processor core loaded from A72 core running Linux through remoteproc and add few peripheral support (Interrupts, Gpio, UART, Timers) for TDA4VM. The Cortex R5 processor cores are built to provide deeply embedded real-time and safety-critical systems. Adding Zephyr RTOS support for R5 cores in TDA4VM will be very helpful for the Users.. Programming Languages: Assembly Programming, C.