yalecyu/crnn.caffe

crnn.caffe

• 👋 Hi, I’m @yalecyu
• 👀 I’m interested in ...
• 🌱 I’m currently learning ...
• 💞️ I’m looking to collaborate on ...
• 📫 How to reach me ...

crnn.caffe's Issues

训练你的代码出现维度不匹配

loss_layer.cpp:19: check failed:bottom[0]->shape(0)==bottom[0]->shape(0)(13 vs. 523) the data and label should have the same first dimension.
*check failure stack trace
请问这是什么原因？

你好，想问下和warpctc-caffe的区别

想问下你的项目下的https://github.com/yalecyu/crnn.caffe/tree/master/src 和https://github.com/xmfbit/warpctc-caffe/tree/master/src 有区别吗？

ContinuationIndicator出自哪里？

您好！请问网络里的ContinuationIndicator是啥，出自哪里参考文献？

测试结果的问题

你好，非常感谢您共享了您的代码，对我有很大的帮助，我在测试的时候有点困惑。我的训练集有150张图片。我的测试图片中的字符是‘UTC4558E’, 可是预测出来的结果是‘UUUTTTCC444555588EE’，为什么会预测出来这么多字符呢？如何去掉这些重复的字符呢？

make fail with protobuf/stubs/common.h not found

I get the following error:

(base) -bash-4.2$ make
CXX .build_release/src/caffe/proto/caffe.pb.cc
In file included from .build_release/src/caffe/proto/caffe.pb.cc:4:0:
.build_release/src/caffe/proto/caffe.pb.h:9:42: fatal error: google/protobuf/stubs/common.h: No such file or directory
 #include <google/protobuf/stubs/common.h>
                                          ^
compilation terminated.
make: *** [.build_release/src/caffe/proto/caffe.pb.o] Error 1

How can I train the network with different image sizes?

Hello,
I am training the network for text recognition, but the same size 128px for all word-instance image is not good because the word doesn't always have the same length.

But I don't know how to change it inside the net, could you please give me some help?

accuracy is 0

我生成了5w的样本，4w做训练，1w做测试，accuracy为0，可能的原因是什么呢，还有就是你的accuracy和ctc loss分别是什么呢

Failed at Make. cuda_compile_generated_ctc_loss_layer.cu.o' failed

Error while make in cmake fashion. Complains

cuda_compile_generated_ctc_loss_layer.cu.o' failed

cmake Log

-- The C compiler identification is GNU 5.4.0
-- The CXX compiler identification is GNU 5.4.0
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working CXX compiler: /usr/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Looking for pthread.h
-- Looking for pthread.h - found
-- Looking for pthread_create
-- Looking for pthread_create - not found
-- Looking for pthread_create in pthreads
-- Looking for pthread_create in pthreads - not found
-- Looking for pthread_create in pthread
-- Looking for pthread_create in pthread - found
-- Found Threads: TRUE  
-- Boost version: 1.58.0
-- Found the following Boost libraries:
--   system
--   thread
--   filesystem
--   chrono
--   date_time
--   atomic
-- Found GFlags: /usr/include  
-- Found gflags  (include: /usr/include, library: /usr/lib/x86_64-linux-gnu/libgflags.so)
-- Found Glog: /usr/include  
-- Found glog    (include: /usr/include, library: /usr/lib/x86_64-linux-gnu/libglog.so)
-- Found Protobuf: /usr/lib/x86_64-linux-gnu/libprotobuf.so  
-- Found PROTOBUF Compiler: /usr/bin/protoc
-- Found HDF5: /usr/lib/x86_64-linux-gnu/hdf5/serial/lib/libhdf5_hl.so;/usr/lib/x86_64-linux-gnu/hdf5/serial/lib/libhdf5.so;/usr/lib/x86_64-linux-gnu/libpthread.so;/usr/lib/x86_64-linux-gnu/libsz.so;/usr/lib/x86_64-linux-gnu/libz.so;/usr/lib/x86_64-linux-gnu/libdl.so;/usr/lib/x86_64-linux-gnu/libm.so (found version "1.8.16") 
-- Found LMDB: /usr/include  
-- Found lmdb    (include: /usr/include, library: /usr/lib/x86_64-linux-gnu/liblmdb.so)
-- Found LevelDB: /usr/include  
-- Found LevelDB (include: /usr/include, library: /usr/lib/x86_64-linux-gnu/libleveldb.so)
-- Found Snappy: /usr/include  
-- Found Snappy  (include: /usr/include, library: /usr/lib/x86_64-linux-gnu/libsnappy.so)
-- CUDA detected: 8.0
-- Found cuDNN: ver. 5.1.10 found (include: /usr/include, library: /usr/lib/x86_64-linux-gnu/libcudnn.so)
-- Automatic GPU detection failed. Building for all known architectures.
-- Added CUDA NVCC flags for: sm_20 sm_21 sm_30 sm_35 sm_50 sm_60 sm_61
-- OpenCV found (/usr/share/OpenCV)
-- Found Atlas: /usr/include  
-- Found Atlas (include: /usr/include, library: /usr/lib/libatlas.so)
-- Found PythonInterp: /usr/bin/python2.7 (found suitable version "2.7.12", minimum required is "2.7") 
-- Found PythonLibs: /usr/lib/x86_64-linux-gnu/libpython2.7.so (found suitable version "2.7.12", minimum required is "2.7") 
-- Found NumPy: /usr/local/lib/python2.7/dist-packages/numpy/core/include (found suitable version "1.13.1", minimum required is "1.7.1") 
-- NumPy ver. 1.13.1 found (include: /usr/local/lib/python2.7/dist-packages/numpy/core/include)
-- Boost version: 1.58.0
-- Found the following Boost libraries:
--   python
-- Could NOT find Doxygen (missing:  DOXYGEN_EXECUTABLE) 
-- Found Git: /usr/bin/git (found version "2.7.4") 
-- 
-- ******************* Caffe Configuration Summary *******************
-- General:
--   Version           :   1.0.0-rc4
--   Git               :   55a9938-dirty
--   System            :   Linux
--   C++ compiler      :   /usr/bin/c++
--   Release CXX flags :   -O3 -DNDEBUG -fPIC -Wall -Wno-sign-compare -Wno-uninitialized
--   Debug CXX flags   :   -g -fPIC -Wall -Wno-sign-compare -Wno-uninitialized
--   Build type        :   Release
-- 
--   BUILD_SHARED_LIBS :   ON
--   BUILD_python      :   ON
--   BUILD_matlab      :   OFF
--   BUILD_docs        :   ON
--   CPU_ONLY          :   OFF
--   USE_OPENCV        :   ON
--   USE_LEVELDB       :   ON
--   USE_LMDB          :   ON
--   USE_NCCL          :   OFF
--   ALLOW_LMDB_NOLOCK :   OFF
-- 
-- Dependencies:
--   BLAS              :   Yes (Atlas)
--   Boost             :   Yes (ver. 1.58)
--   glog              :   Yes
--   gflags            :   Yes
--   protobuf          :   Yes (ver. 2.6.1)
--   lmdb              :   Yes (ver. 0.9.17)
--   LevelDB           :   Yes (ver. 1.18)
--   Snappy            :   Yes (ver. 1.1.3)
--   OpenCV            :   Yes (ver. 2.4.9.1)
--   CUDA              :   Yes (ver. 8.0)
-- 
-- NVIDIA CUDA:
--   Target GPU(s)     :   Auto
--   GPU arch(s)       :   sm_20 sm_21 sm_30 sm_35 sm_50 sm_60 sm_61
--   cuDNN             :   Yes (ver. 5.1.10)
-- 
-- Python:
--   Interpreter       :   /usr/bin/python2.7 (ver. 2.7.12)
--   Libraries         :   /usr/lib/x86_64-linux-gnu/libpython2.7.so (ver 2.7.12)
--   NumPy             :   /usr/local/lib/python2.7/dist-packages/numpy/core/include (ver 1.13.1)
-- 
-- Documentaion:
--   Doxygen           :   No
--   config_file       :   
-- 
-- Install:
--   Install path      :   /home/crnn.caffe/build/install
-- 
-- Configuring done
-- Generating done
-- Build files have been written to: /home/crnn.caffe/build

Make Log

[  1%] Running C++/Python protocol buffer compiler on /home/crnn.caffe/src/caffe/proto/caffe.proto
Scanning dependencies of target proto
[  1%] Building CXX object src/caffe/CMakeFiles/proto.dir/__/__/include/caffe/proto/caffe.pb.cc.o
[  1%] Linking CXX static library ../../lib/libproto.a
[  1%] Built target proto
[  1%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_crop_layer.cu.o
[  1%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/solvers/cuda_compile_generated_nesterov_solver.cu.o
[  1%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/solvers/cuda_compile_generated_sgd_solver.cu.o
[  1%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_hdf5_output_layer.cu.o
[  1%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/solvers/cuda_compile_generated_adagrad_solver.cu.o
[  2%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/solvers/cuda_compile_generated_adam_solver.cu.o
[  4%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_im2col_layer.cu.o
[  4%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/3rdparty/cuda_compile_generated_ctc_entrypoint.cu.o
[  4%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_exp_layer.cu.o
[  4%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_embed_layer.cu.o
[  4%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/3rdparty/cuda_compile_generated_reduce.cu.o
[  4%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/util/cuda_compile_generated_im2col.cu.o
[  5%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/solvers/cuda_compile_generated_adadelta_solver.cu.o
[  5%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/solvers/cuda_compile_generated_rmsprop_solver.cu.o
[  5%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_scale_layer.cu.o
[  5%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_conv_layer.cu.o
[  6%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_cudnn_relu_layer.cu.o
[  8%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_permute_layer.cu.o
[  9%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_log_layer.cu.o
[  9%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_contrastive_loss_layer.cu.o
[ 10%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_cudnn_conv_layer.cu.o
[ 13%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_continuation_indicator_layer.cu.o
[ 13%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_absval_layer.cu.o
[ 13%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_ctc_loss_layer.cu.o
[ 13%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_mvn_layer.cu.o
[ 13%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/util/cuda_compile_generated_math_functions.cu.o
[ 13%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_tanh_layer.cu.o
[ 14%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_recurrent_layer.cu.o
[ 14%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_softmax_layer.cu.o
[ 14%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_filter_layer.cu.o
[ 14%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_prelu_layer.cu.o
[ 14%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_bias_layer.cu.o
[ 14%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_threshold_layer.cu.o
[ 14%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_lrn_layer.cu.o
[ 14%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_relu_layer.cu.o
[ 17%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_silence_layer.cu.o
[ 16%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_dropout_layer.cu.o
[ 17%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_batch_norm_layer.cu.o
[ 18%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_bnll_layer.cu.o
[ 18%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_sigmoid_layer.cu.o
[ 20%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_cudnn_tanh_layer.cu.o
[ 20%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_eltwise_layer.cu.o
[ 20%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_cudnn_softmax_layer.cu.o
[ 20%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_sigmoid_cross_entropy_loss_layer.cu.o
[ 20%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_cudnn_lrn_layer.cu.o
[ 21%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_power_layer.cu.o
[ 22%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_batch_reindex_layer.cu.o
[ 22%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_split_layer.cu.o
[ 22%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_cudnn_pooling_layer.cu.o
[ 22%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_softmax_loss_layer.cu.o
[ 22%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_cudnn_sigmoid_layer.cu.o
[ 24%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_concat_layer.cu.o
[ 27%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_tile_layer.cu.o
[ 28%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_elu_layer.cu.o
[ 28%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_cudnn_lcn_layer.cu.o
[ 28%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_euclidean_loss_layer.cu.o
[ 28%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_pooling_layer.cu.o
[ 28%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_reduction_layer.cu.o
[ 29%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_lstm_unit_layer.cu.o
[ 29%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_hdf5_data_layer.cu.o
[ 29%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_deconv_layer.cu.o
[ 31%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_inner_product_layer.cu.o
[ 31%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_base_data_layer.cu.o
[ 31%] Building NVCC (Device) object src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_slice_layer.cu.o
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
In file included from /usr/include/c++/5/tuple:35:0,
                 from /home/crnn.caffe/include/caffe/3rdparty/detail/cpu_ctc.cuh:3,
                 from /home/crnn.caffe/src/caffe/3rdparty/ctc_entrypoint.cu:6:
/usr/include/c++/5/bits/c++0x_warning.h:32:2: error: #error This file requires compiler and library support for the ISO C++ 2011 standard. This support must be enabled with the -std=c++11 or -std=gnu++11 compiler options.
 #error This file requires compiler and library support \
  ^
CMake Error at cuda_compile_generated_ctc_entrypoint.cu.o.cmake:207 (message):
  Error generating
  /home/crnn.caffe/build/src/caffe/CMakeFiles/cuda_compile.dir/3rdparty/./cuda_compile_generated_ctc_entrypoint.cu.o


src/caffe/CMakeFiles/caffe.dir/build.make:105: recipe for target 'src/caffe/CMakeFiles/cuda_compile.dir/3rdparty/cuda_compile_generated_ctc_entrypoint.cu.o' failed
make[2]: *** [src/caffe/CMakeFiles/cuda_compile.dir/3rdparty/cuda_compile_generated_ctc_entrypoint.cu.o] Error 1
make[2]: *** Waiting for unfinished jobs....
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
/home/crnn.caffe/src/caffe/3rdparty/reduce.cu(44): error: identifier "__shfl_down" is undefined
          detected during:
            instantiation of "T CTAReduce<NT, T, Rop>::reduce(int, T, CTAReduce<NT, T, Rop>::Storage &, int, Rop) [with NT=128, T=float, Rop=ctc_helper::add<float, float>]" 
(76): here
            instantiation of "void reduce_rows<NT,Iop,Rop,T>(Iop, Rop, const T *, T *, int, int) [with NT=128, Iop=ctc_helper::negate<float, float>, Rop=ctc_helper::add<float, float>, T=float]" 
(124): here
            instantiation of "void ReduceHelper::impl(Iof, Rof, const T *, T *, int, int, __nv_bool, cudaStream_t) [with T=float, Iof=ctc_helper::negate<float, float>, Rof=ctc_helper::add<float, float>]" 
(139): here
            instantiation of "ctcStatus_t reduce(Iof, Rof, const T *, T *, int, int, __nv_bool, cudaStream_t) [with T=float, Iof=ctc_helper::negate<float, float>, Rof=ctc_helper::add<float, float>]" 
(149): here

/home/crnn.caffe/src/caffe/3rdparty/reduce.cu(44): error: identifier "__shfl_down" is undefined
          detected during:
            instantiation of "T CTAReduce<NT, T, Rop>::reduce(int, T, CTAReduce<NT, T, Rop>::Storage &, int, Rop) [with NT=128, T=float, Rop=ctc_helper::maximum<float, float>]" 
(76): here
            instantiation of "void reduce_rows<NT,Iop,Rop,T>(Iop, Rop, const T *, T *, int, int) [with NT=128, Iop=ctc_helper::identity<float, float>, Rop=ctc_helper::maximum<float, float>, T=float]" 
(124): here
            instantiation of "void ReduceHelper::impl(Iof, Rof, const T *, T *, int, int, __nv_bool, cudaStream_t) [with T=float, Iof=ctc_helper::identity<float, float>, Rof=ctc_helper::maximum<float, float>]" 
(139): here
            instantiation of "ctcStatus_t reduce(Iof, Rof, const T *, T *, int, int, __nv_bool, cudaStream_t) [with T=float, Iof=ctc_helper::identity<float, float>, Rof=ctc_helper::maximum<float, float>]" 
(157): here

2 errors detected in the compilation of "/tmp/tmpxft_00006cfc_00000000-18_reduce.compute_20.cpp1.ii".
CMake Error at cuda_compile_generated_reduce.cu.o.cmake:266 (message):
  Error generating file
  /home/crnn.caffe/build/src/caffe/CMakeFiles/cuda_compile.dir/3rdparty/./cuda_compile_generated_reduce.cu.o


src/caffe/CMakeFiles/caffe.dir/build.make:112: recipe for target 'src/caffe/CMakeFiles/cuda_compile.dir/3rdparty/cuda_compile_generated_reduce.cu.o' failed
make[2]: *** [src/caffe/CMakeFiles/cuda_compile.dir/3rdparty/cuda_compile_generated_reduce.cu.o] Error 1
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
/home/crnn.caffe/src/caffe/layers/ctc_loss_layer.cu(14): error: explicit type is missing ("int" assumed)

/home/crnn.caffe/src/caffe/layers/ctc_loss_layer.cu(14): error: type name is not allowed

/home/crnn.caffe/src/caffe/layers/ctc_loss_layer.cu(14): error: expected a ";"

/home/crnn.caffe/src/caffe/layers/ctc_loss_layer.cu(15): error: expression must have class type

/home/crnn.caffe/src/caffe/layers/ctc_loss_layer.cu(16): error: expression must have class type

/home/crnn.caffe/src/caffe/layers/ctc_loss_layer.cu(17): error: expression must have class type

/home/crnn.caffe/src/caffe/layers/ctc_loss_layer.cu(26): error: no suitable constructor exists to convert from "int" to "ctcOptions"

/home/crnn.caffe/src/caffe/layers/ctc_loss_layer.cu(37): error: namespace "std" has no member "accumulate"

/home/crnn.caffe/src/caffe/layers/ctc_loss_layer.cu(41): error: expression must have class type

/home/crnn.caffe/src/caffe/layers/ctc_loss_layer.cu(32): error: no suitable constructor exists to convert from "int" to "ctcOptions"
          detected during instantiation of "void caffe::CtcLossLayer<Dtype>::Forward_gpu(const std::vector<caffe::Blob<Dtype> *, std::allocator<caffe::Blob<Dtype> *>> &, const std::vector<caffe::Blob<Dtype> *, std::allocator<caffe::Blob<Dtype> *>> &) [with Dtype=float]" 
(61): here

10 errors detected in the compilation of "/tmp/tmpxft_000071b4_00000000-17_ctc_loss_layer.compute_61.cpp1.ii".
CMake Error at cuda_compile_generated_ctc_loss_layer.cu.o.cmake:266 (message):
  Error generating file
  /home/crnn.caffe/build/src/caffe/CMakeFiles/cuda_compile.dir/layers/./cuda_compile_generated_ctc_loss_layer.cu.o


src/caffe/CMakeFiles/caffe.dir/build.make:364: recipe for target 'src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_ctc_loss_layer.cu.o' failed
make[2]: *** [src/caffe/CMakeFiles/cuda_compile.dir/layers/cuda_compile_generated_ctc_loss_layer.cu.o] Error 1
CMakeFiles/Makefile2:272: recipe for target 'src/caffe/CMakeFiles/caffe.dir/all' failed
make[1]: *** [src/caffe/CMakeFiles/caffe.dir/all] Error 2
Makefile:127: recipe for target 'all' failed
make: *** [all] Error 2

CNN结构替换成denseNet时遇到的问题

您好，谢谢您的无私分享。我将crnn中的cnn结构替换成densenet结构时，遇到了一些问题。

1. solver文件设置test_initialization: true，TEST阶段的ctcloss=nan，accuarcy=0，一直到训练结束。
   我打印训练日志看了下，TEST阶段连个LSTM层的输出均为nan，且bn层的参数值也不正常。
   

2. 设置test_initialization: false时，1的问题就解决了，但是在模型训练好之后，利用模型对测试集进行评估时，准确率在45%左右；实际训练中训练集准确率=1， 验证集准确率为98%；我又用模型对验证集的准确率进行了统计，准确率也在55%左右。

这个问题一直困扰我很久，如果你也碰到了类似问题，请问有没有什么解决办法

where did you set the relationship between "-" and blank label ?

In your project, you set "10" as blank label, consequently, the "10" in test result is replaced as "-". Now I set "37" as blank label, but the "37" in test result is not replaced but all "10" are replaced as "-". Is there any way to change "-" ?

On my MacBookPro, error: no member named 'accumulate' in namespace

CXX src/caffe/layers/ctc_loss_layer.cpp
src/caffe/layers/ctc_loss_layer.cpp:84:23: error: no member named 'accumulate' in namespace 'std'
    Dtype loss = std::accumulate(cost, cost + mini_batch, Dtype(0));
                 ~~~~~^
1 error generated.
make: *** [.build_release/src/caffe/layers/ctc_loss_layer.o] Error 1

关于BN层的参数问题

我测试的时候deploy中设置use_global_stats: false结果就正常，设置为true就不对，可是我看网上说测试的时候应该设置为true，训练prototxt里TEST的时候也是设置为true，请问你有过这样的问题么？为何你给出的deploy里这样设置？

Why not use bidirectional LSTM?

I found that in the origin CRNN implementation, the author used 2 layers of BLSTM, but i only found 2 layers of LSTM in your project. Would you please provide some help for that implementation?

hello,麻烦咨询下我这边有个维度不匹配的情况

您好，我这边不知道什么情况，按照您前两步。第一步生成了100000个png图片，但是第二步生成h5文件的时候，只生成一个train_0.h5,然后我就把trainging.lsit的剩下三个给删除了，然后开始训练网络，好像Reshape的时候，维度不匹配，我也没改prototxt文件，想咨询下，这个问题。。。

F0809 07:18:53.336143 3514 reshape_layer.cpp:87] Check failed: top[0]->count() == bottom[0]->count() (6426624 vs. 8568832) output count must match input count
*** Check failure stack trace: ***
@ 0x7f0d887ff5cd google::LogMessage::Fail()
@ 0x7f0d88801433 google::LogMessage::SendToLog()
@ 0x7f0d887ff15b google::LogMessage::Flush()
@ 0x7f0d88801e1e google::LogMessageFatal::~LogMessageFatal()
@ 0x7f0d88ca7c8f caffe::ReshapeLayer<>::Reshape()

Wrong accuracy when I change test_iter

Good day!
I decided to update your solver.prototxt. So I changed test_iter from 50 to 300 and after that I see that accuracy is lower 6%. Why?

Why only apply batch norm in TEST

I don't understand that part and when i try on VGG Synth dataset, the validation performance is very bad with the BN...

图片是不是必须这个格式，多一点边框都不行吗？

Simple Chinese words recognition

Hi @yalecyu

I am Nic from NVIDIA Shanghai site, I read your crnn caffe code, it's quite interesting.
I have a question about your model: Can it be able to recognize Simple Chinese words if training with Chinese dataset ?
If you are interested in this topic, we can discuss privately([email protected]).
Thanks.

关于模型测试输出的问题

1、请问，对图片进行测试，测试的结果是每一帧对应n个状态的输出，每一帧选取最大的概率进行输出，你是这样做的吗？
2、这个输出的概率代表的含义是什么的，输出每个状态的概率？还是在上一帧输出为某个结果的情况下的条件概率呢？
3、有没有实现过输出结果序列的top3，即概率排列在前几的几条输出路径呢

your model doesn't match to your deploy.txt

when run"./build/examples/cpp_recognition/recognition.bin data/captcha/1.png examples/crnn/deploy.prototxt examples/crnn/model/crnn_captcha.caffemodel",error show "Cannot copy param 0 weights from layer 'conv0'; shape mismatch. Source param shape is 64 3 3 3 (1728); target param shape is 64 1 3 3 (576). To learn this layer's parameters from scratch rather than copying from a saved net, rename the layer."

check failed: registry.count(type)==1(0 vs 1) unknown layer type:

我在VS2015环境下，测试recognition.cpp函数，执行到net_.reset(new Net(model_file, caffe::TEST));时，提示：
I1028 11:15:27.573096 14076 layer_factory.cpp:58] Creating layer data
F1028 11:15:27.574095 14076 layer_factory.cpp:62] Check failed: registry.count(type) == 1 (0 vs. 1) Unknown layer type: Input (known types: Convolution, Eltwise, LRN, Pooling, Power, Python, ReLU, Sigmoid, Softmax, Split, TanH)
*** Check failure stack trace: ***
请问是哪里配置出了问题？

example core dump

./build/examples/cpp_recognition/recognition.bin data/captcha/1.png examples/crnn/deploy.prototxt examples/crnn//model/crnn_captcha.caffemodel
F0222 11:25:31.748042 4965 net.cpp:759] Cannot copy param 0 weights from layer 'conv0'; shape mismatch. Source param shape is 64 3 3 3 (1728); target param shape is 64 1 3 3 (576). To learn this layer's parameters from scratch rather than copying from a saved net, rename the layer.
*** Check failure stack trace: ***
@ 0x7f4f249735cd google::LogMessage::Fail()
@ 0x7f4f24975433 google::LogMessage::SendToLog()
@ 0x7f4f2497315b google::LogMessage::Flush()
@ 0x7f4f24975e1e google::LogMessageFatal::~LogMessageFatal()
@ 0x7f4f24d28a93 caffe::Net<>::CopyTrainedLayersFrom()
@ 0x7f4f24d2f5e5 caffe::Net<>::CopyTrainedLayersFromBinaryProto()
@ 0x7f4f24d2f67e caffe::Net<>::CopyTrainedLayersFrom()
@ 0x40426c Classifier::Classifier()
@ 0x40302e main
@ 0x7f4f1f325830 __libc_start_main
@ 0x403459 _start
@ (nil) (unknown)
Aborted (core dumped)

标签问题

你这标签一定要固定长度5吗？

test_accuracy和模型实际测试的值不同

你好，我想请问下，为什么训练时我的test_accuarcy很高，但是我用训练好的模型对相同的测试集进行测试时，识别效果却很低？是模型过拟合的问题吗

Windows build error

Hi, I try to do a windows build on this project. But, I get the following error when run testing.

Hope can get some help
Thank you

测试例子出现问题

你好，我在make之后执行
./build/examples/cpp_recognition/recognition.bin` data/captcha/1.png examples/crnn/deploy.prototxt examples/crnn/model/crnn_captcha.caffemodel进行测试，但出现了“can not copy param 0 weights from layer 'conv0';shape mismatch.source param shape is 64 3 3 3(1728);target param shape is 64 1 33 (576).......”这个错误，是不是ptototxt 和caffemodel不匹配的问题？麻烦指点一下，多谢~

How to train the 36 Alpha by "0~9 + A~Z"?

Hi @yalecyu ,
Thanks for sharing your code! I have trained the number of 09, and the test result is very effective.
But I can not train the 36 Alpha by "09 + AZ"!
The maximum length of my character is 19, of which there are 09 digits or A~Z characters in it.
How can I train 36 Alpha?
I modified the generate_train_id.py to meet the 36 words, The code I modified is as follows:

#########################################################################
#!/usr/bin/env python

coding=utf-8

import pdb
import os
import numpy as np
from multiprocessing import Process
import sys
sys.path.insert(0,'python')
import caffe
import h5py

CAFFE_ROOT = os.getcwd() # assume you are in $CAFFE_ROOT$ dir
img_path = os.path.join(CAFFE_ROOT, 'data/AlphaNum/train/')
IMAGE_WIDTH, IMAGE_HEIGHT = 128, 32
LABEL_SEQ_LEN = 19

captcha images list

images = filter(lambda x: os.path.splitext(x)[1] == '.jpg', os.listdir(img_path))

print '[+] total image number: {}'.format(len(images))

np.random.shuffle(images)

def write_image_info_into_hdf5(file_name, images, phase):
total_size = len(images)
print '[+] total image for {0} is {1}'.format(file_name, len(images))

single_size = 500
groups = total_size / single_size
if total_size % single_size:
    groups += 1
def process(file_name, images):

    #####################Alpha support################
    Alpha = ['0','1','2','3','4','5','6','7','8','9',\
             'A','B','C','D','E','F','G','H','I','J',\
             'K','L','M','N','O','P','Q','R','S','T',\
             'U','V','W','X','Y','Z']
    ##################################################

    img_data = np.zeros((len(images), 3, IMAGE_HEIGHT, IMAGE_WIDTH), dtype = np.float32)
    label_seq = 10*np.ones((len(images), LABEL_SEQ_LEN), dtype = np.float32)
    for i, image in enumerate(images):
        img_name = os.path.splitext(image)[0]
        newNumStr = img_name[0:]
        
        ############Alpha support##############
        numbers_str = range(len(newNumStr))
        for i in range(0, len(newNumStr)):
            numbers_str[i] = Alpha.index(newNumStr[i])
        #######################################

        numbers = np.array(map(lambda x: float(x), numbers_str))
        label_seq[i, :len(numbers)] = numbers
        img = caffe.io.load_image(os.path.join(img_path, image))
        img = caffe.io.resize(img, (IMAGE_HEIGHT, IMAGE_WIDTH, 3))
        img = np.transpose(img, (2, 0, 1))
        img_data[i] = img
        """
        if (i+1) % 100 == 0:
            print '[+] name: {}'.format(image)
            print '[+] number: {}'.format(','.join(map(lambda x: str(x), numbers)))
            print '[+] label: {}'.format(','.join(map(lambda x: str(x), label_seq[i])))
        """
    with h5py.File(file_name, 'w') as f:
        f.create_dataset('data', data = img_data)
        f.create_dataset('label', data = label_seq)
with open(file_name, 'w') as f:
    workspace = os.path.split(file_name)[0]
    process_pool = []
    for g in xrange(groups):
        h5_file_name = os.path.join(workspace, '%s_%d.h5' %(phase, g))
        f.write(h5_file_name + '\n')
        start_idx = g*single_size
        end_idx = start_idx + single_size
        if g == groups - 1:
            end_idx = len(images)
        p = Process(target = process, args = (h5_file_name, images[start_idx:end_idx]))
        p.start()
        process_pool.append(p)
    for p in process_pool:
        p.join()

trainning_size = 2789 # number of images for trainning
trainning_images = images[:trainning_size]

write_image_info_into_hdf5(os.path.join('data/AlphaNum/train_datasets/', 'trainning.list'), trainning_images, 'train')
##############################################################################

But I don't know if I've modified it correctly. Could you give me some suggestions?
And what form should "crnn.prototxt" be changed into?

Any of your suggestions will be geate helpful to me!
I am looking forward to your reply!
Thank you!

Sincerely,
Samylee

你好，训练自己数据的时候准确率和loss都为0

你好，我在用自己数据训练的时候，出现准确率和loss都为0的情况，我的训练数据是1050张，测试数据是50张，如下，我一开始以为是batch太大了，所以我改成64，但还是这样，请问是为什么
I0505 20:36:57.910152 15093 solver.cpp:331] Iteration 0, Testing net (#0)
I0505 20:36:58.462410 15093 solver.cpp:398] Test net output #0: accuracy = 0
I0505 20:36:58.462445 15093 solver.cpp:398] Test net output #1: ctc_loss = 0.0889562 (* 1 = 0.0889562 loss)
I0505 20:36:58.503353 15093 solver.cpp:219] Iteration 0 (-2.35495e-09 iter/s, 0.593967s/10 iters), loss = 0
I0505 20:36:58.503383 15093 solver.cpp:238] Train net output #0: accuracy = 0
I0505 20:36:58.503389 15093 solver.cpp:238] Train net output #1: ctc_loss = 0 (* 1 = 0 loss)
I0505 20:36:58.503413 15093 sgd_solver.cpp:105] Iteration 0, lr = 0.001
I0505 20:36:58.832192 15093 solver.cpp:219] Iteration 10 (30.4131 iter/s, 0.328806s/10 iters), loss = 0
I0505 20:36:58.832212 15093 solver.cpp:238] Train net output #0: accuracy = 0
I0505 20:36:58.832217 15093 solver.cpp:238] Train net output #1: ctc_loss = 0 (* 1 = 0 loss)
I0505 20:36:58.832235 15093 sgd_solver.cpp:105] Iteration 10, lr = 0.001
I0505 20:36:59.160102 15093 solver.cpp:219] Iteration 20 (30.4983 iter/s, 0.327887s/10 iters), loss = 0
I0505 20:36:59.160135 15093 solver.cpp:238] Train net output #0: accuracy = 0
I0505 20:36:59.160140 15093 solver.cpp:238] Train net output #1: ctc_loss = 0 (* 1 = 0 loss)
I0505 20:36:59.160143 15093 sgd_solver.cpp:105] Iteration 20, lr = 0.001
I0505 20:36:59.487865 15093 solver.cpp:219] Iteration 30 (30.5132 iter/s, 0.327727s/10 iters), loss = 0
I0505 20:36:59.487886 15093 solver.cpp:238] Train net output #0: accuracy = 0
I0505 20:36:59.487892 15093 solver.cpp:238] Train net output #1: ctc_loss = 0 (* 1 = 0 loss)
I0505 20:36:59.487895 15093 sgd_solver.cpp:105] Iteration 30, lr = 0.001
I0505 20:36:59.814749 15093 solver.cpp:219] Iteration 40 (30.5943 iter/s, 0.326858s/10 iters), loss = 0
I0505 20:36:59.814783 15093 solver.cpp:238] Train net output #0: accuracy = 0
I0505 20:36:59.814790 15093 solver.cpp:238] Train net output #1: ctc_loss = 0 (* 1 = 0 loss)
I0505 20:36:59.814839 15093 sgd_solver.cpp:105] Iteration 40, lr = 0.001
I0505 20:37:00.145546 15093 solver.cpp:219] Iteration 50 (30.2332 iter/s, 0.330763s/10 iters), loss = 0
I0505 20:37:00.145583 15093 solver.cpp:238] Train net output #0: accuracy = 0
I0505 20:37:00.145589 15093 solver.cpp:238] Train net output #1: ctc_loss = 0 (* 1 = 0 loss)
I0505 20:37:00.145593 15093 sgd_solver.cpp:105] Iteration 50, lr = 0.001
I0505 20:37:00.474901 15093 solver.cpp:219] Iteration 60 (30.3663 iter/s, 0.329312s/10 iters), loss = 0
I0505 20:37:00.474920 15093 solver.cpp:238] Train net output #0: accuracy = 0
I0505 20:37:00.474926 15093 solver.cpp:238] Train net output #1: ctc_loss = 0 (* 1 = 0 loss)
I0505 20:37:00.474931 15093 sgd_solver.cpp:105] Iteration 60, lr = 0.001
I0505 20:37:00.800871 15093 solver.cpp:219] Iteration 70 (30.6796 iter/s, 0.32595s/10 iters), loss = 0
I0505 20:37:00.800900 15093 solver.cpp:238] Train net output #0: accuracy = 0
I0505 20:37:00.800906 15093 solver.cpp:238] Train net output #1: ctc_loss = 0 (* 1 = 0 loss)
I0505 20:37:00.800909 15093 sgd_solver.cpp:105] Iteration 70, lr = 0.001
I0505 20:37:01.128326 15093 solver.cpp:219] Iteration 80 (30.5412 iter/s, 0.327426s/10 iters), loss = 0
I0505 20:37:01.128342 15093 solver.cpp:238] Train net output #0: accuracy = 0
I0505 20:37:01.128348 15093 solver.cpp:238] Train net output #1: ctc_loss = 0 (* 1 = 0 loss)
I0505 20:37:01.128367 15093 sgd_solver.cpp:105] Iteration 80, lr = 0.001
I0505 20:37:01.455250 15093 solver.cpp:219] Iteration 90 (30.5905 iter/s, 0.326899s/10 iters), loss = 0
I0505 20:37:01.455288 15093 solver.cpp:238] Train net output #0: accuracy = 0
I0505 20:37:01.455296 15093 solver.cpp:238] Train net output #1: ctc_loss = 0 (* 1 = 0 loss)
I0505 20:37:01.455301 15093 sgd_solver.cpp:105] Iteration 90, lr = 0.001
I0505 20:37:01.783249 15093 solver.cpp:219] Iteration 100 (30.4918 iter/s, 0.327956s/10 iters), loss = 0
I0505 20:37:01.783283 15093 solver.cpp:238] Train net output #0: accuracy = 0
I0505 20:37:01.783289 15093 solver.cpp:238] Train net output #1: ctc_loss = 0 (* 1 = 0 loss)
I0505 20:37:01.783293 15093 sgd_solver.cpp:105] Iteration 100, lr = 0.001

When I compile the project, an error "src/caffe/3rdparty/reduce.cu(44): error: identifier "__shfl_down" is undefined"

When I compile the project, an error "src/caffe/3rdparty/reduce.cu(44): error: identifier "__shfl_down" is undefined", how to solve it ?

将CNN换成densenet结构后，BN层设置问题

我将前面CNN结构替换成densenet结构，BN层按照train-false, test-true这样设置参数，ctcloss=nan，请问这个情况，有遇到过吗

测试结果出现了很大问题

我训练模型达到了0.96的accurary，用./build/examples/cpp_recognition/recognition.bin data/testimage/1.png examples/crnn/deploy.prototxt examples/crnn/model/crnn_captcha_iter_20000.caffemodel测试的时候无论用哪张图，结果只能显示74 74 74 1 74 4 74 74 1 74 74 4 74 74 1 74 74 74 74 74 74 74 74 74 74 74 74 74 74 74 74 74这样的数字串，请问是什么原因造成的呢

What is the difference between this caffe and original caffe？

Can you please tell me what layers are different from original caffe？
I think this addition add CtcLoss layer and ContinuationIndicator layer. Is that right?
Where is the cpp file of the added layers?
The original caffe has a lstm_layer.cpp.I notice that you have change that cpp.So can you tell me what else you have changed?
THX in advance :)

Alphabets in label's questions

@yalecyu Hi,bro.
There are some troubles about my own data.Because there are some alphabets in my data like 'a'、'b'、'c',I can not just make it into np.array.And I don't know how to change the "generate_dataset.py".Can you give me a template.
Looking forward to your reply

自制数据集，过拟合的问题

你好，我在训练自制数据集的时候，发现网络特别容易就过拟合，我的训练样本在13000张左右，训练集的准确率很快就到了100，但是测试集的准确率只有70左右，不知道如何解决这个问题呢？