Stack Overflow
Stack Overflow

Questions tagged [caffe]

Caffe is a fast deep learning framework. It supports CPU and GPU processing. Caffe is released under the BSD 2-Clause license. It is based on C++ but provides APIs for Python and Matlab.

Caffe is a fast framework, developed by Berkeley AI Research (BAIR)/ The Berkeley Vision and Learning Center (BVLC). Its architecture is based on pure and and supports and processing without the need to adapt the code for one of the processing units. Caffe provides and interfaces. Caffe is released under the BSD 2-Clause license.

Caffe users share popular models for different tasks with all kinds of architectures and data in the Model Zoo. These models are learned and applied for problems ranging from simple regression, to large-scale visual classification, to Siamese networks for image similarity, to speech and robotics applications.

Useful Links:

2834 questions
1
vote
2 answers

Converting Caffe caffe::Datum to OpenCV cv::Mat in C++

I'm doing some debugging and so I'm dumping image files to look at the predictions and transformations. I can create a caffe::Datum from cv::Mat: cv::Mat matrix; // ... initialize matrix caffe::Datum datum; caffe::CVMatToDatum(matrix, &datum) but…
empty
  • 5,194
  • 3
  • 32
  • 58
1
vote
1 answer

how to understand the upsample layer of SegNet

I notice that there are upsample layers in SegNet,its own images are 480*360,when I want to use my images(565*584),I meet the following error: I0929 03:58:06.238135 22750 net.cpp:368] upsample4 -> pool4_D I0929 03:58:06.238142 22750 net.cpp:120]…
StalkerMuse
  • 1,001
  • 3
  • 11
  • 22
1
vote
0 answers

googlenet testing technique multi-crop with 144 images in caffe

In the googlenet's paper: we resized the image to 4 scales where the shorter dimension (height or width) is 256, 288, 320 and 352 respectively, take the left, center and right square of these resized images. This gives a more accuracy in testing.…
李德健
  • 51
  • 3
1
vote
1 answer

how to prepare gray image data for caffe input

I have read lots of references to prepare image data for caffe input but have not been able to solve my problem yet!! I have put .jpg image files in 2 seperate folders: folder train: containing 100,000 image files of size 227*227 folder val:…
user6726469
  • 231
  • 1
  • 3
  • 14
1
vote
1 answer

undefined reference to symbol '_ULx86_64_step' libunwind.so.8: error adding symbols: DSO missing from command line

I'm trying to compile Caffe. make all -j4 worked fine. But making the tests is giving the following error. Debian GNU/Linux 8 (jessie) 64-bit g++ --version g++ (Debian 4.9.2-10) 4.9.2 Gives this error on make test -j4 caffe$ make test…
muglikar
  • 146
  • 2
  • 18
1
vote
1 answer

NuGet error while building caffe on Windows with visual studio 2013

I'm trying to build Caffe on Windows in order to use it on Python (by Import caffe) for my Deep Learning project, but I came across a problem while building the Caffe.sln file with Visual Studio 2013, following instructions from this video :…
filipyoo
  • 347
  • 1
  • 7
  • 15
1
vote
1 answer

Is Caffe::Net Multithread-safe?

Is Caffe::Net multithread-safe? Can you have several threads each with its own Caffe::Net in test mode? It appears to me, looking at the code, that Caffe assumes only one instance of Net which launches multiple threads to access multiple GPUs.
empty
  • 5,194
  • 3
  • 32
  • 58
1
vote
2 answers

Running simple C++ programm using Caffe

I am trying to run my simple program with caffe and i got a error #include #include #include using namespace std; using namespace caffe; int main(int argc, char** argv) { shared_ptr > net_; …
Ivan Gunko
  • 13
  • 3
1
vote
2 answers

Creating lmdb file in the right way

I'm trying to create an lmdb file that contains all of my database images (in order to train CNN). This is my 'test code', that I took from here: import numpy as np import lmdb import caffe import cv2 import glob N = 18 # Let's pretend this is…
roishik
  • 515
  • 2
  • 9
  • 19
1
vote
1 answer

Making hdf5 data with 2d mat file in MATLAB

I am trying to do binary classfication in caffe. I have 900 training samples and each sample is a 30-by-15 .mat file and label is a 900-by-1 .mat file(+1,-1). Now, my question is how can I make hdf5 data to feed the caffe network. I know how to do…
Mohaiminul Islam
  • 29
  • 4
1
vote
1 answer

SegNet - architecture & image size

Im trying to train my own dataset on SegNet (with caffe), I prepared the dataset same as segnet tutorial. when I try to run the train, it shows me this error: I0915 05:08:08.281393 45783 layer_factory.hpp:74] Creating layer data I0915…
Z.Kal
  • 426
  • 4
  • 18
1
vote
1 answer

Failed to find HDF5 dataset data - Single Label Regression Using Caffe and HDF5 data

I was using @shai 's code for creating my hdf5 file which is available here: Test labels for regression caffe, float not allowed? My data is grayscale images (1000 images for a start) and label is one dimensional float numbers So, I modified his…
magneto
  • 97
  • 6
1
vote
1 answer

How to use leveldb and What kind of dataLayer I can use in pycaffe interface?

I tried to make train/val.prototxt using leveldb by caffe python interface: layer { name: "cifar" type: "Data" top: "data" top: "label" data_param { source:…
youngwan lee
  • 93
  • 2
  • 7
1
vote
1 answer

why does not clEnequeMapBuffer map to original pointer, OpenCL, Caffe

Assume that a CPU pointer(cpu_ptr_) already exists, then I create a buffer for gpu(cl_gpu_mem_). The problem is that when I map the gpu buffer to a cpu pointer(mapped_ptr), the mapped_ptr is not equal to the original pointer (cpu_ptr_), which causes…
Jake Ren
  • 13
  • 1
  • 6
1
vote
0 answers

Control caffe display while training

I am using caffe to train a deep network, and have set the display at 200 iterations in my solver prototext file. However, instead of getting the loss and accuracy, I am getting a large number of lines of the form solver.cpp:245] Train net output…
quantumcoder
  • 33
  • 1
  • 1
  • 7
1 2 3
…
99
100