SoonminHwang / caffe-1.0

Caffe: a fast open framework for deep learning.

Updates

Several layers are added for compatibility to deeplab-v2, etc.

• layer/seg_accuracy_layer (.cpp, .hpp)

• layer/interp layer (.cpp, .hpp)

• util/interp (.cpp, .cu)

• util/confusion_matrix (.cpp, .hpp)

• common.cuh: Add follows (for CUDA 8.0)

    #if !defined(__CUDA_ARCH__) || __CUDA_ARCH__ >= 600
    #else
    // CUDA: atomicAdd is not defined for doubles
    static __inline__ __device__ double atomicAdd(double *address, double val) {
      unsigned long long int* address_as_ull = (unsigned long long int*)address;
      unsigned long long int old = *address_as_ull, assumed;
      if (val==0.0)
        return __longlong_as_double(old);
      do {
        assumed = old;
        old = atomicCAS(address_as_ull, assumed, __double_as_longlong(val +__longlong_as_double(assumed)));
      } while (assumed != old);
      return __longlong_as_double(old);
    }
    #endif
  

• layer/BN_layer (.cpp, .cu, .hpp)

• layer/image_seg_data_layer (.cpp, .hpp)

• Updates: utils/io, data transformer, layer/base data layer, Makefile (Add CXXFLAGS += --std=c++11 in Line 12)

• Common things: PREFETCH_COUNT -> prefetch_.size()

• Enhanced Image Seg Data Layer with data augmentation, which is based on: @twtygqyy's caffe-augmentation.

  • Usage:

    In training phase,

      layer {
      	name: "data"
      	type: "ImageSegData"
      	top: "data"
      	top: "mask"
      	include {
      	  phase: TRAIN
      	}
      	transform_param {
      	    mirror: true
      		crop_size: 224
    
      		mean_value: 104.008
      		mean_value: 116.669
      		mean_value: 122.675
      		scale_factors: 0.5
      		scale_factors: 1.0
      		scale_factors: 1.5
      		scale_factors: 2.0
    
      	    contrast_brightness_adjustment: true
      	    smooth_filtering: true
      	    min_side_min: 256
      	    min_side_max: 480				    				    
      	    min_contrast: 0.8
      	    max_contrast: 1.2
      	    max_smooth: 6
      	    apply_probability: 0.5
      	    max_color_shift: 20
      	    debug_params: false
      	}
      	image_data_param {
      	  source: "train_list.txt"
      	  batch_size: 64
      	}
      }
    

    In testing phase,

      layer {
      	name: "data"
      	type: "ImageData"
      	top: "data"
      	top: "label"
      	include {
      	  phase: TEST
      	}
      	transform_param {
      	    mirror: false
      	    min_side: 256
      	    crop_size: 224
      	    mean_value: 104.008
      		mean_value: 116.669
      		mean_value: 122.675
      	}
      	image_data_param {
      	  source: "test_list.txt"
      	  batch_size: 32
      	}
      }
    

---

Caffe

Caffe is a deep learning framework made with expression, speed, and modularity in mind. It is developed by Berkeley AI Research (BAIR)/The Berkeley Vision and Learning Center (BVLC) and community contributors.

Check out the project site for all the details like

• DIY Deep Learning for Vision with Caffe
• Tutorial Documentation
• BAIR reference models and the community model zoo
• Installation instructions

and step-by-step examples.

Please join the caffe-users group or gitter chat to ask questions and talk about methods and models. Framework development discussions and thorough bug reports are collected on Issues.

Happy brewing!

License and Citation

Caffe is released under the BSD 2-Clause license. The BAIR/BVLC reference models are released for unrestricted use.

Please cite Caffe in your publications if it helps your research:

@article{jia2014caffe,
  Author = {Jia, Yangqing and Shelhamer, Evan and Donahue, Jeff and Karayev, Sergey and Long, Jonathan and Girshick, Ross and Guadarrama, Sergio and Darrell, Trevor},
  Journal = {arXiv preprint arXiv:1408.5093},
  Title = {Caffe: Convolutional Architecture for Fast Feature Embedding},
  Year = {2014}
}