Julien Mairal  SoftwareThe software packages below are either written by me, or by my students, when mentioned. SPAMSSPAMS is an optimization toolbox implementing algorithms to address various machine learning and signal processing problems involving dictionary learning and matrix factorization (e.g., NMF, sparse PCA); solving mediumscale sparse decomposition problems with LARS, coordinate descent, OMP, SOMP, proximal methods; solving largescale sparse estimation problems with stochastic optimization; solving structured sparse decomposition problems (e.g., sparse group lasso, treestructured regularization, structured sparsity with overlapping groups). The code was mostly written by me. Interfaces with Python and R were developed by JeanPaul Chieze (Inria). Latest releases for Python3 and R3 were packaged and are maintained by Ghislain Durif (Inria). CKNCmatlabThis is a reimplementation of the convolutional kernel network (CKN) methods introduced in Julien Mairal. EndtoEnd Kernel Learning with Supervised Convolutional Kernel Networks. Adv. NIPS. 2016. This is an almost pure C implementation using directly CUDA and CUDNN, along with a Matlab interface. The software package features both the unsupervised and supervised variants of CKNs and is opensource with a GPLv3 license. FlipFlopFlipFlop is an opensource software, implementing a fast method for de novo transcript discovery and abundance estimation from RNASeq data. It differs from classical approaches such as Cufflinks by simultaneously performing the identification and quantitation tasks using a penalized maximum likelihood approach, which leads to improved precision recall. Other software taking this approach have an exponential complexity in the number of exons of a gene. We use a novel algorithm based on network flow formalism, which gives us a polynomial runtime. In practice, FlipFlop was shown to outperform penalized maximum likelihood based software in terms of speed, and to perform transcript discovery in less than 1/2 second for large genes. FlipFlop 1.0.0 is a user friendly bioconductor R package. It is freely available on the Bioconductor website. The code was written coauthored by Elsa Bernard (Institut Curie), Laurent Jacob (CNRS), and me. CKNseqCKNseq is a software package to model biological sequences with convolutional kernel networks. The current implementation corresponds to the BiorXiv preprint “Predicting Transcription Factor Binding Sites withConvolutional Kernel Networks”. It was written by Dexiong Chen (Inria). LoterThis is a software package for local ancestry inference corresponding to the BiorXiv preprint “Loter: A Software Package to Infer Local Ancestry for a Wide Range of Species”. The package is written and maintained by Thomas DiasAlves (Univ. Grenoble Alpes). StochsThis is the opensource software package corresponding to the NIPS’17 paper “ Stochastic Optimization with Variance Reduction for Infinite Datasets with FiniteSum Structure” for largescale machine learning problems with perturbations. The package is written and maintained by Alberto Bietti (Inria). MODLThis is the opensource software package corresponding to the ICML’16 paper “Dictionary Learning for Massive Matrix Factorization” for hugescale matrix factorization. The package is written and maintained by Arthur Mensch (Inria). This is a highly optimized library that is able to handle matrices of several terabytes on a single workstation. BlitzNetThis is the opensource software package corresponding to the ICCV’17 paper “BlitzNet: A RealTime Deep Network for Scene Understanding”. This is a realtime deep network for object detection and scene segmentation. It is written and maintained by Mikita Dvornik (Inria). CKNThis is the opensource software package corresponding to the paper “Convolutional kernel networks” published at NIPS’14. We are going to release soon a new GPU implementation corresponding to the NIPS’16 paper “Endtoend kernel learning with supervised convolutional kernel networks”. PatchCKNThis is the software package and the new dataset “RomePatches” corresponding to the IJCV paper “Convolutional Patch Representations for Image Retrieval: an Unsupervised Approach”. DolphinThis is the opensource software package corresponding to the paper IEEETSP paper “DOLPHInDictionary Learning for Phase Retrieval”. The package is written and maintained by Andreas Tillmann (TU Darmstadt). Misc
