bnll

include/caffe/vision_layers.hpp

@@ -45,6 +45,25 @@ class ReLULayer : public NeuronLayer<Dtype> {
 };
 
 
+template <typename Dtype>
+class BNLLLayer : public NeuronLayer<Dtype> {
+ public:
+  explicit BNLLLayer(const LayerParameter& param)
+      : NeuronLayer<Dtype>(param) {}
+
+ protected:
+  virtual void Forward_cpu(const vector<Blob<Dtype>*>& bottom,
+      vector<Blob<Dtype>*>* top);
+  virtual void Forward_gpu(const vector<Blob<Dtype>*>& bottom,
+      vector<Blob<Dtype>*>* top);
+
+  virtual Dtype Backward_cpu(const vector<Blob<Dtype>*>& top,
+      const bool propagate_down, vector<Blob<Dtype>*>* bottom);
+  virtual Dtype Backward_gpu(const vector<Blob<Dtype>*>& top,
+      const bool propagate_down, vector<Blob<Dtype>*>* bottom);
+};
+
+
 template <typename Dtype>
 class DropoutLayer : public NeuronLayer<Dtype> {
  public:

src/caffe/layer_factory.cpp

@@ -21,6 +21,8 @@ Layer<Dtype>* GetLayer(const LayerParameter& param) {
   const std::string& type = param.type();
   if (type == "accuracy") {
     return new AccuracyLayer<Dtype>(param);
+  } else if (type == "bnll") {
+    return new BNLLLayer<Dtype>(param);
   } else if (type == "conv") {
     return new ConvolutionLayer<Dtype>(param);
   } else if (type == "data") {

src/caffe/layers/bnll_layer.cu

@@ -0,0 +1,89 @@
+// Copyright 2013 Yangqing Jia
+
+#include "caffe/layer.hpp"
+#include "caffe/vision_layers.hpp"
+#include <algorithm>
+
+using std::max;
+
+namespace caffe {
+
+const float kBNLL_THRESHOLD = 50.;
+
+template <typename Dtype>
+void BNLLLayer<Dtype>::Forward_cpu(const vector<Blob<Dtype>*>& bottom,
+    vector<Blob<Dtype>*>* top) {
+  const Dtype* bottom_data = bottom[0]->cpu_data();
+  Dtype* top_data = (*top)[0]->mutable_cpu_data();
+  const int count = bottom[0]->count();
+  for (int i = 0; i < count; ++i) {
+    top_data[i] = log(1. + exp(min(bottom_data[i], Dtype(kBNLL_THRESHOLD))));
+  }
+}
+
+template <typename Dtype>
+Dtype BNLLLayer<Dtype>::Backward_cpu(const vector<Blob<Dtype>*>& top,
+    const bool propagate_down,
+    vector<Blob<Dtype>*>* bottom) {
+  if (propagate_down) {
+    const Dtype* bottom_data = (*bottom)[0]->cpu_data();
+    const Dtype* top_diff = top[0]->cpu_diff();
+    Dtype* bottom_diff = (*bottom)[0]->mutable_cpu_diff();
+    const int count = (*bottom)[0]->count();
+    for (int i = 0; i < count; ++i) {
+      Dtype expval = exp(min(bottom_data[index], Dtype(kBNLL_THRESHOLD)));
+      bottom_diff[index] = top_diff[index] * expval / (expval + 1.);
+    }
+  }
+  return Dtype(0);
+}
+
+template <typename Dtype>
+__global__ void BNLLForward(const int n, const Dtype* in, Dtype* out) {
+  int index = threadIdx.x + blockIdx.x * blockDim.x;
+  if (index < n) {
+    out[index] = log(1. + exp(min(in[index], Dtype(kBNLL_THRESHOLD)));
+  }
+}
+
+template <typename Dtype>
+void BNLLLayer<Dtype>::Forward_gpu(const vector<Blob<Dtype>*>& bottom,
+    vector<Blob<Dtype>*>* top) {
+  const Dtype* bottom_data = bottom[0]->gpu_data();
+  Dtype* top_data = (*top)[0]->mutable_gpu_data();
+  const int count = bottom[0]->count();
+  BNLLForward<Dtype><<<CAFFE_GET_BLOCKS(count), CAFFE_CUDA_NUM_THREADS>>>(
+      count, bottom_data, top_data);
+  CUDA_POST_KERNEL_CHECK;
+}
+
+template <typename Dtype>
+__global__ void BNLLBackward(const int n, const Dtype* in_diff,
+    const Dtype* in_data, Dtype* out_diff) {
+  int index = threadIdx.x + blockIdx.x * blockDim.x;
+  if (index < n) {
+    Dtype expval = exp(min(in_data[index], Dtype(kBNLL_THRESHOLD)));
+    out_diff[index] = in_diff[index] * expval / (expval + 1.);
+  }
+}
+
+template <typename Dtype>
+Dtype BNLLLayer<Dtype>::Backward_gpu(const vector<Blob<Dtype>*>& top,
+    const bool propagate_down,
+    vector<Blob<Dtype>*>* bottom) {
+  if (propagate_down) {
+    const Dtype* bottom_data = (*bottom)[0]->gpu_data();
+    const Dtype* top_diff = top[0]->gpu_diff();
+    Dtype* bottom_diff = (*bottom)[0]->mutable_gpu_diff();
+    const int count = (*bottom)[0]->count();
+    BNLLBackward<Dtype><<<CAFFE_GET_BLOCKS(count), CAFFE_CUDA_NUM_THREADS>>>(
+        count, top_diff, bottom_data, bottom_diff);
+    CUDA_POST_KERNEL_CHECK;
+  }
+  return Dtype(0);
+}
+
+INSTANTIATE_CLASS(BNLLLayer);
+
+
+}  // namespace caffe