Contribo
ProgramsOrganizationsProjectsProposal StudioResourcesOrbit AIDashboard
Login
Login
Contribov1.0

Your premium companion for discovering, matching, and tracking paid open-source mentorship programs and verified starter issues.

Product

  • Programs Directory
  • Orbit AI
  • User Dashboard

Resources

  • Guides & Resources
  • Proposal Studio

Community

  • GitHub Repository
  • Code of Conduct

Legal

  • Privacy Policy
  • Terms & Conditions
  • Cookie Policy
  • Accessibility Statement

Made with love bySubhranshu Sekhar Dash

© 2026 Contribo. Built for the global open-source community.

PlatformProjects

Explore Projects

Find open source projects actively accepting contributors. Search repositories, filter by program milestones, difficulty tags, or tech stack.

Confused which project to choose or don't know which one you are most suitable to work on?

Use our Orbit AI Matcher to find out! Get instant matching scores based on your developer skills, preferred frameworks, and contribution experience.

Match with Orbit AI
Filters Clear
3179 Projects Found
Sort by

Recurrent Neural Networks for ChainerX

intermediate
NumFOCUS•2019Varies

Ready to Apply?

Convert your selected open-source project into a winning GSoC, LFX, or Outreachy application using Proposal Studio.

Start Proposal
Proposal Status
0% ProgressNot Started
Accepted Proposals
3 SimilarView →
Proposal Guide
Tailored DocsOpen →
Checklist & Review
Readiness AuditCheck →
175h (Medium)

<p>In this project, I propose to add Recurrent Neural Networks to ChainerX. RNNs are a very integral part of deep learning research. ChainerX is faster as compared to other frameworks (eg: Chainer, PyTorch) since its core code is completely implemented in C++. It currently lacks implementations of RNNs. In this project, I will provide C++ implementations of RNN models (LSTM, GRU). This will bring the fast speed of ChainerX to RNN models. I aim to replicate all RNN-related chainer ops in ChainerX. I will also provide sufficient code examples on how to use the RNN models of ChainerX.</p>

pythonjavascriptr+3
Student: Aniket Didolkar
Code
Proposal

Implement Interface for external data source import into Calc

intermediate
LibreOffice•2021Varies175h (Medium)

<p>Currently back end of data provider is implemented and it supports csv,html,xml and base data provider. Many transformations are also available so that user can have a look of data after applied transformation before importing it into calc sheet.The data provider was removed in commit <a href="https://git.libreoffice.org/core/+/24939a5b43b9fe889c3aace027c668666ff61987%5E%21" target="_blank">https://git.libreoffice.org/core/+/24939a5b43b9fe889c3aace027c668666ff61987%5E%21</a> because there were many issues in UI. So the aim is to improve the interface of data providers and solving issues of data transformations to bring back it into production.</p>

pythonandroidjava+2
Student: Tushar Kumar Rai
Code
Proposal

Development of Qutip Functions for analysis of Bosonic/Fermionic Lattices

intermediate
NumFOCUS•2019Varies175h (Medium)

<p>The following functionality is aimed at: Choice of Model (lattice dimensionality, Boson/Fermion, Hubbard, XXZ, Boundary conditions etc.). Non-interacting/Mean Field Models:<br> Band Structure and Eigen-state calculations. (Canonical diagonalization of Fermions/Bosons). Ability to take band structures as inputs directly for subsequent calculations. Visualization of iso-energy surfaces in the band-structure. Berry Curvature (From dynamics/Gauge potential). Chern numbers. Interacting models: Exact diagonalization. Ground state and Entanglement Spectrum calculation with DMRG. Dynamics with td-DMRG. Calculation of Conductivity</p>

pythonjavascriptr+3
Student: Saumya Biswas
Code
Proposal

Improving TensorFlow GAN Library

intermediate
TensorFlow•2021Varies175h (Medium)

<p>Tensorflow GAN is a lightweight library that is used for training and evaluating GAN models. Through this project different variants of GAN models will be implemented and trained using TF-GAN, and will be added as examples to the library. Additionally notebook tutorials for these examples will also be added which can also help users to gain insights into the implementation, training and evaluation process for these models and can help in exploring different useful features of the TF-GAN library. This project is aimed at providing users a more convenient way to understand the benefits of using TF-GANs, encouraging them to make use of the library.</p>

pythonc/c++machine learning+8
Student: Nived P A
Code
Proposal

OCI compatible Kubernetes ontology

Intermediate
•Varies175h (Medium)

Network topologies and graph databases go hand-in-hand. The OpenAPI specifications for Kubernetes provides taxonomy, but augmenting a graph data model with formalized ontologies enables any number of capabilities, one of the more straightforward is the inferencing requisite for natural language processing, and consequently, a human-centric query / response interaction becomes becomes possible. More importantly, more advanced systems can be built when a graph data model of connected systems is upgraded to be a knowledge semantic graph. Deliverables (among other items): - MeshModel capabilities browser - Import/export of MeshModel models and components as OCI images - augmentation of cuelang-based component generator

TBA
Proposal

Representing Kubernetes ontology in MeshModel

Intermediate
•Varies175h (Medium)

Network topologies and graph databases go hand-in-hand. The OpenAPI specifications for Kubernetes provides taxonomy, but augmenting a graph data model with formalized ontologies enables any number of capabilities, one of the more straightforward is the inferencing requisite for natural language processing, and consequently, a human-centric query / response interaction becomes becomes possible. More importantly, more advanced systems can be built when a graph data model of connected systems is upgraded to be a knowledge semantic graph. Deliverables (among other items): - MeshModel capabilities browser - Import/export of MeshModel models and components as OCI images - augmentation of cuelang-based component generator

TBA
Proposal

PictoPy

intermediate
AOSSIE•2024Varies175h (Medium)

This project seeks to create a desktop application for an image gallery, with machine learning models powering its functionality in the background. This will enable features such as object, face, and scene tagging. The user interface will mimic that of a traditional gallery app, with capabilities like expansion and sliding. Data will be stored in lightweight databases utilizing SQLite, while the backend will employ YOLO architecture-driven (and similiar) models. At the end of this project, my aim is to establish a streamlined ML pipeline integrating multiple models. This pipeline will preprocess and store photos, allowing users to search for specific types of photos based on generated tags.

pythonjavascriptswift+8
Student: Bassam Adnan
Code
Proposal

JAX ML Textbook

advanced
TensorFlow•2022Varies350h (Large)

The project will be making contributions to the PyProbML repo helping to build demonstrations and reference implementations for probabilistic machine learning models in Jax. This project will involve multiple discrete contributions of varying size and complexity aiming for a particular focus on projects involving: temporal modelling, latent variable modelling, and Bayesian factor graphs. There is also the potential to further develop the contribution I have made to the repo on restricted Boltzmann machines and contrastive divergence. This could include, for example, implementing a Deep Boltzmann Machine in Jax/Flax, demonstrating pre-training with contrastive divergence, and implementing more advanced forms of contrastive divergence such as persistent contrastive divergence.

pythonc/c++machine learning+8
Student: Giles Harper-Donnelly
Code
Proposal

Finding Exoplanets with Astronomical Observations

advanced
Machine Learning for Science (ML4SCI)•2022Varies350h (Large)

The ultimate purpose of this project is to train image-based deep learning models to detect planets embedded in protoplanetary disks. This will be done using a kinematic analysis of the disk, i.e., observing how the material inside of it is moving. Given the relative sparsity of high-quality kinematic observations of these disks, training data will be a combination of actual observations and synthetic images generated from hydrodynamics simulations. Models will not only be evaluated based on their performance as quantified via typical measures, but they will also be evaluated on their ability to locate planets through the intensity of their activations. A performant model will confidently predict the presence, or lack-thereof, of a planet and precisely highlight it when activated.

pythonmachine learningc+++2
Student: Jason Terry
Code

Torch compile and PyTorch 2.2.0

advanced
DeepChem•2024Varies350h (Large)

This project aims to integrate the torch.compile function released in PyTorch 2.0 into DeepChem's PyTorch models, unlocking significant performance improvements for training and inference. By leveraging advanced JIT compilation techniques used in torch.compile, this project will enable DeepChem users to optimize their models for faster execution without requiring code modifications. Initial benchmarks demonstrate promising speedups for various architectures, and further testing will be conducted to ensure broad compatibility and optimal performance gains across DeepChem's diverse model library. This project will ultimately empower researchers and developers with faster and more efficient tools for scientific discovery and innovation.

pythonnumpypytorch+1
Student: Gautham Krishnan
Code
Proposal

Move2Kube Simplify plugin architecture of m2k (2024 Term 1)

Intermediate
•Varies175h (Medium)

Move2Kube is a command-line tool for automating creation of Infrastructure as code (IaC) artifacts. The tool has a powerful plugin based transformer architecture where developers can write their own custom transformer plugins to fulfil their re-platforming needs. However, concepts like path mappings etc could be simplified for better adoption. Example - writing a Move2Kube custom transformer needs developers to understand various concepts such as path mappings etc, can we reduce this learning overhead by simplifying the Move2Kube architecture? Expected Outcome: - Come up with a simplified alternative design for plugin architecture for M2K - Migrate existing components to support the new design with backward compatibility (good to have).

TBA
Proposal

Simulating Representational Communication in Vervet Monkeys

intermediate
Red Hen Lab•2021Varies175h (Medium)

<p>Vervet monkeys (Cercopithecus aethiops) are said to give acoustically different alarm calls to different predators, evoking contrasting, seemingly adaptive, responses. Since it can be a potentially critical evolutionary feature if proven its worth, there is a need to check how effective are these alarm calls in helping the vervets' overall survival. This project is a predator-prey agent-based modeling attempt to simulate an ecological setup consisting of vervet monkeys as they interact with predators and give alarm calls. Hence the main goal of the simulation is to efficiently model representational communication through an agent-based model and show how the differential meaning of vervet’s alarm calls is helping in their survival.</p>

machine learningopencvhigh performance computing+15
Student: ankiitgupta7
Code
Proposal

EXXA - Denoising Astronomical Observations of Protoplanetary Disks

advanced
Machine Learning for Science (ML4SCI)•2026Varies350h (Large)

Astronomical observations of protoplanetary disks often contain noise that obscures important structures such as rings and gaps that may indicate planet formation. This project proposes to develop a machine learning pipeline for denoising astronomical observations using deep learning techniques including convolutional autoencoders, U-Net architectures, and diffusion models. The project will implement a complete workflow for loading FITS datasets, preprocessing observations, training denoising models, and evaluating performance using PSNR, SSIM, and MSE metrics. The final deliverables will include a reproducible open source pipeline, trained models, and evaluation tools for improving the quality of protoplanetary disk observations.

pythonmachine learningc+++2
Student: Krishan Yadav
Code

Implement RFDiffusion and RFDiffusion All-Atom in DeepChem

advanced
DeepChem•2026Varies350h (Large)

RFDiffusion (Watson et al., Nature 2023) is the state-of-the-art protein backbone diffusion model but DeepChem currently has no protein structure generative model. This project implements RFDiffusion and RFDiffusion All-Atom as native DeepChem TorchModel subclasses, covering a SE(3)-equivariant denoiser with Invariant Point Attention, a RoseTTAFold-style multi-track architecture, SO(3) diffusion for rotations, conditioning mechanisms (motif scaffolding, symmetry, binder design), the all-atom extension for side chains and small molecules, and a full evaluation pipeline. Deliverables include merged PRs for the featurizer, CATH dataset loader, core layers, and the complete model, along with benchmark results and a tutorial notebook.

pythonnumpypytorch+1
Student: chidwipak
Code
Proposal

Transfer Learning Framework

intermediate
Open Chemistry•2019Varies175h (Medium)

<p>Even though chemistry has become a more data driven discipline in the recent years, the amount of data available for training deep learning models is limited when compared to their imaging counterparts (ImageNet, for example). Transfer learning is a strategy that aims to leverage representations learnt, by training deep learning models on larger datasets with available labels, and then using the trained network for fine-tuning on smaller, costly to label datasets.</p> <p>This project is about porting a Transfer Learning Framework, ChemNet into DeepChem. This would involve reproducing the results of the paper and a blog-post, Jupyter notebook detailing its use. A ChemNet API will also be developed to allow any DeepChem TensorGraph model to use this framework.</p>

pythonjavascriptopengl+7
Student: Vignesh Ram Somnath
Code
Proposal

add threats by element for STRIDE/LINDDUN/PLOT4ai/CIA/DIE

intermediate
OWASP Foundation•2024Varies175h (Medium)

Reinstating automated threat suggestion per element was a feature in the threat dragon threat model designing system. Hence, upon migration to Vue.JS framework, not all features where migrated as that would take much effort to be completed. Yet, the project is continuously reinstating old features with new ones to make threat dragon an even better threat model designing system. This specific project is about suggesting threat type automatically for an element upon creation when drawing a threat model. My solution is about tracking our current existing threats for this element. and upon creating a new threat, the algorithm would check the existing threats and with the map it has for matching suggestions, it would pick the threat type suggestion for the next created threat.

pythonjavascriptjava+10
Student: Mohamed El-Bohy
Code

End-to-End Deep Learning Reconstruction for CMS Experiment

advanced
CERN-HSF•2020Varies350h (Large)

<p>Developing End-to-End Deep Learning Models and optimizing them for the Reconstruction of single particles, jets and event topologies of interest in collision during collision experiment. This involves classifying electromagnetic showers of the particles developed during the experiment and their event classification. The code will be further integrated with the CMSSW inference engine. The inference of the model will be integrated with the CMSSW Particle Flow (PF) classes and will be tested on GPUs for faster Computation. This proposal shows the proposed methodology and model(which may be improvised while working on it depending on the performance and new approaches may be used) for the reconstruction for CMS experiment. It also includes timeline and some basic concepts of Physics to familiarize with the CMS experiment at Large Hadron Collider (LHC).</p>

pythonc/c++machine learning+6
Student: Shravan Chaudhari-1
Code

PyTorch-Ignite: Development of a Higher-level API

advanced
NumFOCUS•2021Varies350h (Large)

<p>In this project we are trying to provide a High-level API for the Deep Learning engineers and researchers to help them train their models faster, and provide them many other various features:</p> <ul> <li>Users can train/evaluate/test models with this API.</li> <li>Users can train multiple models/optimizers on multiple datasets.</li> <li>We provide a lot of features like (integration of experiment tracking systems, auto-batching, auto-checkpointing, automatic-mixed-precision, train GANs, Semi-supervised learning, DDP, etc....)</li> <li>We provide examples for using each and every feature in this API.</li> <li>We provide good documentation.</li> </ul>

pythonjavascriptr+3
Student: Ahmed-1
Code
Proposal

Super Rapid Annotator - Multimodal vision tool to annotate videos with LLaVA framework

intermediate
Red Hen Lab•2024Varies175h (Medium)

This proposal outlines the development of an innovative system designed to enhance video annotation capabilities through the integration of a multimodal vision and language model with spatial-temporal analysis. Utilizing the LLaVA-v1.5-13b model with Video Adapter, known for its rapid inference and entity detection in video frames, this system aims to extract annotations automatically and compile them into a user-friendly CSV format. At its core, a Pydantic API extracts boolean values for annotation entities from enhanced language model responses, facilitating the concurrent processing of multiple videos. The entire workflow, from video and JSON input to CSV output, will be accessible via a Gradio interface on the Hugging Face platform, ensuring ease of use and broad accessibility.

machine learningopencvhigh performance computing+15
Student: Manish Kumar Thota
Code

All-in-One System Identification Toolkit for ArduPilot Platforms

intermediate
ArduPilot•2024Varies175h (Medium)

This project intends to extend the system identification capabilities of Ardupilot to include all UAV types supported by Ardupilot, along with a dedicated library for processing, analysing and extracting the vehicle’s dynamic model from flight data. This will hopefully allow researchers to develop better model based guidance and control algorithms. The project would involve creation of lua scripts for Arduplane, which enables actuating all the control actuators in either a close loop or open loop mode. The flight logs will then be processed via a separate dedicated codebase to analyse and validate the data, and identify transfer functions and state space models. Finally, the custom control library can be extended for ArduPlane to allow researchers to experiment with multiple control algorithms.

pythonrosc+++8
Student: Astik Srivastava
Code

Implementing and Visualizing MP-BioPath tool in the FI Viz Cytoscape Plugin

intermediate
Open Genome Informatics•2022Varies175h (Medium)

This project entails the creation of the MP-BioPath tool into the FI Viz Cytoscape Plugin. Biological pathways model the set of actions that molecules in a biological cell can perform. What the MP-BioPath does is model the activity of certain biological pathways, helping to predict the response of the pathways to external perturbances. The MP-Biopath tool already exists, but has not been made accessible through the FI Viz Cytoscape Plugin, which is an app designed to analyze the pathways involved with diseases such as cancer. My job would be to incorporate the MP-BioPath tool into the FI Viz Cytoscape Plugin and develop customized visualizations for showing the result of MP-BioPath. If I have time remaining, I intend to deliver a baseline Tensorflow model which utilizes the MP-BioPath tool to create predictions for pathway activity.

pythonjavanodejs+9
Student: ModeEric
Code

MiniZinc Interface for R

beginner
R project for statistical computing•2020Varies90h (Small)

<p>MiniZinc is a free and open-source constraint modeling language. Constraint satisfaction and discrete optimization problems can be formulated in a high-level modeling language. Models are compiled into an intermediate representation that is understood by a wide range of solvers. MiniZinc itself provides several solvers, for instance GeoCode. The existing packages in R are not powerful enough to solve even mid-sized problems in combinatorial optimization</p> <p>There are implementations of an Interface to MiniZinc in Python like MiniZinc Python and pymzn and JMiniZinc for Java but such an interface does not exist for R.</p> <p>This proposal describes a way to implement a very simple and easy to use interface for R that will help R users to solve optimization problems that can't be solved with R currently.</p>

cc++r+4
Student: Akshit Achara
Code
Proposal

Exercise using Deep Learning for Visual Detection

intermediate
JdeRobot - Universidad Rey Juan Carlos•2021Varies175h (Medium)

<p>The main goal of this project is to build a new template and a backend architecture running on a docker container for visual detection exercises. More specifically, it revolves around building a template for visual detection exercises(e.g. Human pose estimation or traffic sign recognition), which must be solved by providing a trained model for the particular exercise. In this way the user would have to upload a trained model through a user-friendly web template, the model must match the defined set of inputs(e.g. Single or multiple images) and outputs(e.g. 2D coordinates). This would help replace the previous requirements of manually coding the behavior of the robot. Finally, we would also be building new exercises in the Robotics Academy which make use of this new paradigm.</p>

pythonc/c++ros+4
Student: Shashwat Dalakoti
Code

Improve/Extend weighted transfer rules module

intermediate
Apertium•2019Varies175h (Medium)

<p>Ambiguous patterns are ones that more than one transfer rule could be applied to. Apertium resolves this ambiguity by applying the left-to-right longest match (LRLM) rule, and that is not adequate with all the word/s that follow that pattern/s. To improve this resolution, a new module was introduced to assign weights to these transfer rules for the word/s that follow the ambiguous pattern, and this is done by training a corpus to generate maximum entropy models (models with weighted rules), then these models are used to choose the best (highest weighted) transfer rule to apply.</p> <p>The weighted transfer rules module was built to apply only chunker transfer rules (patterns of words). And this project is to improve it by modify some of the methods used and then to extend it to be applied to interchunk and postchunk transfer rules (patterns of chunks) too.</p>

pythonc++xml+4
Student: Aboelhamd Aly
Code
Proposal
Showing 1825 - 1848 of 3179 Projects
7576777879
Proposal
Proposal
Proposal
Proposal
Proposal
Proposal
Proposal
Proposal