Keras is an open source, cross platform, and user friendly neural network library written in Python. It is capable of running on top of TensorFlow, Microsoft Cognitive Toolkit, R, Theano, and PlaidML.
📝 "End-to-end Deep Learning of Optimization Heuristics" (🥇 PACT'17 Best Paper)
Simulation based Soft Continuum Robot Control via Reinforcement Learning
Classification of flowers from Oxford Flowers 102 dataset
Batik motif classification (5 classes) using VGG16 transfer learning
Dropout vs. batch normalization: effect on accuracy, training and inference times - code for the paper
Binary convnet for food image recognizion (Keras tensorflow)
iJazzARTIST: Intelligent Jazz Accompanist for Real-Time human-computer Interactive muSic improvisaTion ( B.Sc. Thesis )
University final year project: Deep neural networks for classification of multispectral images. Includes dissertation, research Jupyter Notebooks, and code for TensorFlow CNN models.
Adversarial Machine Learning Attacks in Scaled Self-Driving Cars is the topic of my Ms thesis research at the University of Tartu
Machine Learning Phase Transitions
Numerical Linear Algebra meets Machine Learning - PSE Project
Generating Trump tweets with deep learning. Spoiler: the results suck. The new GPT models do *much* better.
This repository contains all the code and material related to the computer security course project regarding the subject of adversarial machine learning.
This repository contains the code for the paper "Icon Generation with Conditional GANs".
Files regarding the matura thesis (2019) of Luis Wirth at the Gymnasium Oberwil
Repository for the course MO444 - Machine Learning and Pattern Recognition
Image classification with Keras on the Belgium Traffic Sign Dataset
An unsupervised learning approach to uniform sampling.
Calculate bone age using the deep learning technique.
Created by François Chollet
Released March 27, 2015