20 best open source convolution projects.

We have collection of more than 1 Million open source products ranging from Enterprise product to small libraries in all platforms. We aggregate information from all open source repositories. Search and find the best for your needs. Check out projects section.

Displaying 1 to 20 from 20 results

imaging - Simple Go image processing package

Package imaging provides basic image manipulation functions (resize, rotate, flip, crop, etc.). This package is based on the standard Go image package and works best along with it.Image manipulation functions provided by the package take any image type that implements image.Image interface as an input, and return a new image of *image.NRGBA type (32bit RGBA colors, not premultiplied by alpha).

image image-processing resize crop rotate convolution blur go-library

VkFFT - Vulkan/CUDA/HIP/OpenCL Fast Fourier Transform library

VkFFT is an efficient GPU-accelerated multidimensional Fast Fourier Transform library for Vulkan/CUDA/HIP/OpenCL projects. VkFFT aims to provide the community with an open-source alternative to Nvidia's cuFFT library while achieving better performance. VkFFT is written in C language and supports Vulkan, CUDA, HIP and OpenCL as backends. Vulkan version: Include the vkFFT.h file and glslang compiler. Provide the library with correctly chosen VKFFT_BACKEND definition (VKFFT_BACKEND=0 for Vulkan). Sample CMakeLists.txt file configures project based on Vulkan_FFT.cpp file, which contains examples on how to use VkFFT to perform FFT, iFFT and convolution calculations, use zero padding, multiple feature/batch convolutions, C2C FFTs of big systems, R2C/C2R transforms, R2R DCT-II, III and IV, double precision FFTs, half precision FFTs. For single and double precision, Vulkan 1.0 is required. For half precision, Vulkan 1.1 is required.

hpc vulkan opencl cuda convolution fft hip dct r2c r2r vulkan-fft c2r

MNN is a blazing fast, lightweight deep learning framework, battle-tested by business-critical use cases in Alibaba

MNN is a highly efficient and lightweight deep learning framework. It supports inference and training of deep learning models, and has industry leading performance for inference and training on-device. At present, MNN has been integrated in more than 20 apps of Alibaba Inc, such as Taobao, Tmall, Youku, Dingtalk, Xianyu and etc., covering more than 70 usage scenarios such as live broadcast, short video capture, search recommendation, product searching by image, interactive marketing, equity distribution, security risk control. In addition, MNN is also used on embedded devices, such as IoT.

machine-learning arm deep-neural-networks deep-learning vulkan ml convolution embedded-devices mnn winograd-algorithm artificial-intelligence

TuSimple-DUC - Understanding Convolution for Semantic Segmentation

Python

by Panqu Wang, Pengfei Chen, Ye Yuan, Ding Liu, Zehua Huang, Xiaodi Hou, and Garrison Cottrell. This repository is for Understanding Convolution for Semantic Segmentation (WACV 2018), which achieved state-of-the-art result on the CityScapes, PASCAL VOC 2012, and Kitti Road benchmark.

semantic-segmentation deep-neural-networks mxnet cityscapes convolution deep-learning

Conditional-PixelCNN-decoder - Tensorflow implementation of Gated Conditional Pixel Convolutional Neural Network

Python

This is a Tensorflow implementation of Conditional Image Generation with PixelCNN Decoders which introduces the Gated PixelCNN model based on PixelCNN architecture originally mentioned in Pixel Recurrent Neural Networks. The model can be conditioned on latent representation of labels or images to generate images accordingly. Images can also be modelled unconditionally. It can also act as a powerful decoder and can replace deconvolution (transposed convolution) in Autoencoders and GANs. A detailed summary of the paper can be found here. The gating accounts for remembering the context and model more complex interactions, like in LSTM. The network stack on the left is the Vertical stack that takes care of blind spots that occure while convolution due to the masking layer (Refer the Pixel RNN paper to know more about masking). Use of residual connection significantly improves the model performance.

deep-learning generative-algorithm paper convolution deepmind tensorflow

SOTA-Py - A discrete-time Python-based solver for the Stochastic On-Time Arrival routing problem

Python

SOTA-Py is a Python-based solver for the policy- and path-based "SOTA" problems, using the algorithm(s) described in Tractable Pathfinding for the Stochastic On-Time Arrival Problem (also in the corresponding arXiv preprint) and previous works referenced therein. What is the SOTA problem? Read on...

routing-algorithm routing transportation transportation-planning transportation-network transportation-problem arrival stochastic-models convolution stochastic-dynamic-programming a-star dijkstra-shortest-path shortest-path-routing-algorithm shortest-pathfinding-algorithm pathfinding-algorithm shortest-path-problem reliable-routing

ConvNetSharp - Deep Learning in C#

CSharp

Started initially as C# port of ConvNetJS. You can use ConvNetSharp to train and evaluate convolutional neural networks (CNN). You must have CUDA version 8 and Cudnn version 6.0 (April 27, 2017) installed. Cudnn bin path should be referenced in the PATH environment variable.

convolution neural-network machine-learning

ndarray-fft - FFT for ndarrays

Javascript

A fast Fourier transform implementation for ndarrays. You can use this to do image processing operations on big, higher dimensional typed arrays in JavaScript.Executes a fast Fourier transform on the complex valued array x/y.

ndarray fft fourier transform convolution bluestein radix 2 image volume filter signal

etl - Expression Templates Library (ETL) with GPU support

ETL is a header only library for C++ that provides vector and matrix classes with support for Expression Templates to perform very efficient operations on them. At this time, the library support compile-time sized matrix and vector and runtime-sized matrix and vector with all element-wise operations implemented. It also supports 1D and 2D convolution, matrix multiplication (naive algorithm and Strassen) and FFT.

c-plus-plus performance cpu gpu cpp14 cpp11 matrix expression-template cpp convolution

deep-scite - :rowboat: A simple recommendation engine (by way of convolutions and embeddings) written in TensorFlow

HTML

DeepScite takes in papers (titles, abstracts) and emits recommendations on whether or not they should be scited by the particular users whose data we've used for training (in the case of this repo, it is me). As output, it also gives a "goodness" score for each word; when this number is high, it has contributed strongly to the paper being (recommended) for sciting, when it is negative, it has contributed strongly to the paper not being recommended.

deep-learning embeddings convolution tensorboard tensorflow

glsl-gaussian - glsl-gaussian is a shader generator for WebGL, to generate a gaussian blur of an input texture

Javascript

glsl-gaussian is a shader generator for WebGL, to generate a gaussian blur of an input texture. See glsl-gaussian-live-demo.js, glsl-gaussian-demo.js for usage.

webgl gl graphics computer-graphics opengl glsl data shader image-processing dsp convolution kernel filter blur summed-area-table box-blur gaussian downsample downsampling subsample subsampling scaling mipmap

glsl-sat - glsl-sat is a shader generator for WebGL, to generate a summed-area-table texture of an input texture

Javascript

glsl-sat is a shader generator for WebGL, to generate a summed-area-table texture of an input texture. See glsl-sat-demo.js for usage.

webgl gl graphics computer-graphics opengl glsl data shader image-processing dsp convolution kernel filter blur summed-area-table

massiv - Efficient Haskell Arrays featuring Parallel computation

Haskell

massiv is a Haskell library for array manipulation. Performance is one of its main goals, thus it is able to run effortlessly almost all operations in parallel as well as sequentially. The name for this library comes from the Russian word Massiv (Масси́в), which means an Array.

haskell array arrays stencil multidimensional-arrays convolution parallel-computing parallel-processing

SkimCaffe - Caffe for Sparse Convolutional Neural Network

SkimCaffe has been only tested with bvlc_reference_caffenet, bvlc_googlenet, and resnet, and there could be places where things do not work if you use other networks. Please let us know if you encounter such issues and share .prototxt of the network you are using. We will try our best to support it as well. Eventually, our sparse CNN implementation should be general enough to handle all kinds of networks. We assume you have a recent Intel compiler and MKL installed. Tested environments: Intel compiler version 15.0.3.187 or newer. boost 1.59.0 . MKL 2017 or newer to use MKL-DNN. Direct sparse convolution and sparse fully-connected layers is only tested for AlexNet and GoogLeNet.

caffe convolution sparsity intel winograd pruning

PYNQ-DL - Xilinx Deep Learning IP

VHDL

After the setup, new Jupyter notebooks will be added under the pynqDL folder, ready to try out, no additional steps are needed.

darius convolutional-neural-networks resize-images convolution resize

pySpeechRev - This python code performs an efficient speech reverberation starting from a dataset of close-talking speech signals and a collection of acoustic impulse responses

Python

This python code performs an efficient speech reverberation starting from a dataset of close-talking speech signals and a collection of acoustic impulse responses. where x[n] is the clean signal and * is the convolutional operator.

speech-recognition distant-speech-recognition speech-reverberation data-contamination impulse-response convolution