The Center for Open Software Innovation (COSI) was established in 2018 within the LTCI Lab.
Its objectives are:
- To present the open source projects created and co-created at LTCI;
- To promote open development practices in computing as an important part of the open science movement;
- to promote open source software and models in academia and in general;
raise awareness among researchers and the public and provide relevant information to interested parties;
- to provide the best possible support for open source software development communities (OSS).
A page of Useful Resources is also available.
Here is the list of open source projects to which LTCI members contribute.
The names shown are those of the LTCI members who contribute to the project, but they may not be the creators or managers of the project, their name is simply listed as the contact point for the project within Télécom Paris.
This list is under development and you can help to extend or correct it by sending an email to firstname.lastname@example.org.
Multimedia, image, sound, video
Bayesian Collaborative Denoising
Collaborative Bayesian image denoising for Monte Carlo rendering algorithms.
Light field representation method for improving the importance sampling of 3D Monte Carlo renderings.
Automatic and generic measures of verbal alignment in a dialogue, based on sequential pattern mining.
Platform covering various aspects of multimedia, from formats such as MP4 or MPEG-2 TS to presentation technologies: graphics, animation and interactivity.
Vector graphics editor for GNU/Linux, Windows and MacOS X that allows creating artistic and technical illustrations such as cartoons, clip art, logos...
Coordinate system with quadruple and triple cages, intermediate between high quality coarse meshing and volume diffusion by spatial coordinates.
High-precision interactive selection system for similar 3D surface regions.
Extreme approximation of the representation of 3D shapes according to a spherical interpolation.
Method of compressing 3D image vectors for a distributed Monte Carlo rendering engine.
Data Science and Artificial Intelligence
AMIE is a system for identifying Horn's associative rules in knowledge bases.
Combinatorial Creativity Tool
Combinatorial creativity combines existing concepts in a new way to produce new concepts.
Evolife is a genetic algorithm in which individuals represent variants of solutions to a problem.
Computational models for the automated analysis of cohesion in small groups interactions.
Platform for data stream extraction with a collection of statistical learning algorithms and evaluation tools.
Path search library for the Rust programming language.
Statistical learning platform for distributed flows that allows the development of algorithms without the complexity of processing engines.
Python library for statistical learning: random forests, logistic regressions, classification algorithms and support vector machines.
Python library for multi-output/multi-label data and data streams that complements scikit-learn.
Python library for the analysis of large graphs such as social networks, web graphs and relational data.
StreamDM allows data mining within data streams with Spark Streaming.
One of the first major public knowledge bases, Yago integrates information from Wikidata and schema.org to give them intelligence and context.
Library to use Ada's BSD sockets (TCP, UDP and multicast), even in a proprietary application.
Interpreter of Forth written in Ada.
Allows to use the readline library of the Ada language, in projects like aforth, with a nice interactive interface.
OpenLISP is an open source implementation of the LISP protocol running in the kernel of the FreeBSD operating system.
Forth compiler on Unix or Linux using gforth 0.6.x which generates code for the Microchip PIC 16F87x and 16F88 family of microcontrollers.
Forth compiler for PIC 18Fxxx microcontrollers, with native compilation, 16-bit cells and optimized generated code.
Networks and communicating objects
This overlay network brings flexibility to the OPC-UA business middleware, the main communication protocol for Industry 4.0.
Support platform for the design of critical embedded real-time systems.
Toolbox for embedded systems dedicated to the edition of UML and SysML diagrams, as well as their simulation and formal verification.