Projects
-
Squeeze-and-excitation Networks: TensorFlow implementation of Squeeze-and-excitation blocks on different convolutional neural networks such as ResNet-18, ResNeXt-29 and InceptionV3 using CIFAR-10, CIFAR-100 and Tiny ImageNet. [Report].
-
Image classification using merged CNNs - Kaggle challenge: a use of two merged pre-trained convolutional neural networks (ResNet-50V2 and InceptionV3) to get a top score in a bottle classification Kaggle challenge. The model predicts how much water or soda a bottle contains.
-
Analysis of pro-level competitive matches in Counter-Strike Global Offensive: a data visualization and exploration of matches in a first person shooter game using R and D3.
-
Graph animation guide with gg animate and After Effects: a guide on graph animations with a workflow involving gganimate and Adobe After Effects.
-
Linear time densest subgraph approximation: analysis of very large graphs requires approximation algorithms to efficiently retrieve information. A linear time of a 2-approximation algorithm was implemented in Python and C++ [Report].
-
Decentralized Task-distribution in Peer-to-Peer Networks: Netlogo implementation of a multi-agent system that achieves decentralized task-distribution in peer-to-peer networks. A set of agents aim to distribute a number of tasks of different types equally among themselves by coordination via a communication network. [Report (French)]
-
3D Maze using the Java 3D api: a 3D maze using Java3D with a dynamic camera, adaptive mini-map and auto solver functionality.
-
Green Guy Project: innovation project - online platform that allows users to promote ecology through geo-located events. (Team of 7 students working on a community-based project)
-
Forum shooter Android version - Android port of an Unreal engine 4 first person shooter: a solo game project made to promote a career fair event organized by students at Telecom Paris. Changes had to be made to port the game to Android.
-
Forum shooter - Unreal engine 4 first person shooter: a solo UE4 game project made to promote a career fair event organized by students at Telecom Paris.