first commit of ResRNN project

Circle-RNN.md

@@ -0,0 +1,13 @@
+# ResRNN
+Implementation of a two path ResRNN architecture (Circle-RNN was invented).
+
+### This repository is an implementation of our IPMI 2017 paper:
+Xue W., Nachum I.B., Pandey S., Warrington J., Leung S., Li S. (2017) Direct Estimation of Regional Wall Thicknesses via Residual Recurrent Neural Network. In: Information Processing in Medical Imaging. IPMI 2017. Lecture Notes in Computer Science, vol 10265. Springer.
+
+https://link.springer.com/chapter/10.1007/978-3-319-59050-9_40 or https://arxiv.org/pdf/1705.09728.pdf
+
+### ON Circle-RNN
+
+In caffe.proto, add the following line to message RecurrentParameter {}
+
+optional uint32 depth = 6 [default = 0];

Circle-RNN.md~

@@ -0,0 +1,5 @@
+
+
+In caffe.proto, add the following line to message RecurrentParameter {}
+
+optional uint32 depth = 6 [default = 0];

Circle-RNN/Circle-RNN.md

@@ -0,0 +1,2 @@
+Circle-RNN
+

Circle-RNN/circle_lstm_layer.cpp

@@ -0,0 +1,271 @@
+#include <string>
+#include <vector>
+
+#include "caffe/blob.hpp"
+#include "caffe/common.hpp"
+#include "caffe/filler.hpp"
+#include "caffe/layer.hpp"
+#include "caffe/layers/circle_lstm_layer.hpp"
+#include "caffe/util/math_functions.hpp"
+
+namespace caffe {
+
+template <typename Dtype>
+void CircleLSTMLayer<Dtype>::RecurrentInputBlobNames(vector<string>* names) const {
+  names->resize(2);
+  (*names)[0] = "h_0";
+  (*names)[1] = "c_0";
+}
+
+template <typename Dtype>
+void CircleLSTMLayer<Dtype>::RecurrentOutputBlobNames(vector<string>* names) const {
+  names->resize(2);
+  int depth = this->layer_param_.recurrent_param().depth();
+  if (depth==0) {
+      depth = this->T_;
+  }
+  (*names)[0] = "h_" + format_int(depth*this->T_);
+  (*names)[1] = "c_T";
+}
+
+template <typename Dtype>
+void CircleLSTMLayer<Dtype>::RecurrentInputShapes(vector<BlobShape>* shapes) const {
+  const int num_output = this->layer_param_.recurrent_param().num_output();
+  const int num_blobs = 2;
+  shapes->resize(num_blobs);
+  for (int i = 0; i < num_blobs; ++i) {
+    (*shapes)[i].Clear();
+    (*shapes)[i].add_dim(1);  // change from a single timestep to all timestep  @xwolfs 2016/10/25
+    (*shapes)[i].add_dim(this->N_);
+    (*shapes)[i].add_dim(num_output);
+  }
+}
+
+template <typename Dtype>
+void CircleLSTMLayer<Dtype>::OutputBlobNames(vector<string>* names) const {
+  names->resize(1);
+  (*names)[0] = "h";
+}
+
+template <typename Dtype>
+void CircleLSTMLayer<Dtype>::FillUnrolledNet(NetParameter* net_param) const {
+  const int num_output = this->layer_param_.recurrent_param().num_output();
+  CHECK_GT(num_output, 0) << "num_output must be positive";
+  const FillerParameter& weight_filler =
+      this->layer_param_.recurrent_param().weight_filler();
+  const FillerParameter& bias_filler =
+      this->layer_param_.recurrent_param().bias_filler();
+
+  // Add generic LayerParameter's (without bottoms/tops) of layer types we'll
+  // use to save redundant code.
+  LayerParameter hidden_param;
+  hidden_param.set_type("InnerProduct");
+  hidden_param.mutable_inner_product_param()->set_num_output(num_output * 4);
+  hidden_param.mutable_inner_product_param()->set_bias_term(false);
+  hidden_param.mutable_inner_product_param()->set_axis(2);
+  hidden_param.mutable_inner_product_param()->
+      mutable_weight_filler()->CopyFrom(weight_filler);
+
+  LayerParameter biased_hidden_param(hidden_param);
+  biased_hidden_param.mutable_inner_product_param()->set_bias_term(true);
+  biased_hidden_param.mutable_inner_product_param()->
+      mutable_bias_filler()->CopyFrom(bias_filler);
+
+  LayerParameter sum_param;
+  sum_param.set_type("Eltwise");
+  sum_param.mutable_eltwise_param()->set_operation(
+      EltwiseParameter_EltwiseOp_SUM);
+
+  LayerParameter scale_param;
+  scale_param.set_type("Scale");
+  scale_param.mutable_scale_param()->set_axis(0);
+
+  LayerParameter slice_param;
+  slice_param.set_type("Slice");
+  slice_param.mutable_slice_param()->set_axis(0);
+
+  LayerParameter split_param;
+  split_param.set_type("Split");
+
+  vector<BlobShape> input_shapes;
+  RecurrentInputShapes(&input_shapes);
+  CHECK_EQ(2, input_shapes.size());
+
+  LayerParameter* input_layer_param = net_param->add_layer();
+  input_layer_param->set_type("Input");
+  InputParameter* input_param = input_layer_param->mutable_input_param();
+
+  input_layer_param->add_top("c_0");
+  input_param->add_shape()->CopyFrom(input_shapes[0]);
+
+  input_layer_param->add_top("h_0");
+  input_param->add_shape()->CopyFrom(input_shapes[1]);
+
+  LayerParameter* cont_slice_param = net_param->add_layer();
+  cont_slice_param->CopyFrom(slice_param);
+  cont_slice_param->set_name("cont_slice");
+  cont_slice_param->add_bottom("cont");
+  cont_slice_param->mutable_slice_param()->set_axis(0);
+
+  // Add layer to transform all timesteps of x to the hidden state dimension.
+  //     W_xc_x = W_xc * x + b_c
+  {
+    LayerParameter* x_transform_param = net_param->add_layer();
+    x_transform_param->CopyFrom(biased_hidden_param);
+    x_transform_param->set_name("x_transform");
+    x_transform_param->add_param()->set_name("W_xc");
+    x_transform_param->add_param()->set_name("b_c");
+    x_transform_param->add_bottom("x");
+    x_transform_param->add_top("W_xc_x");
+    x_transform_param->add_propagate_down(true);
+  }
+
+  if (this->static_input_) {
+    // Add layer to transform x_static to the gate dimension.
+    //     W_xc_x_static = W_xc_static * x_static
+    LayerParameter* x_static_transform_param = net_param->add_layer();
+    x_static_transform_param->CopyFrom(hidden_param);
+    x_static_transform_param->mutable_inner_product_param()->set_axis(1);
+    x_static_transform_param->set_name("W_xc_x_static");
+    x_static_transform_param->add_param()->set_name("W_xc_static");
+    x_static_transform_param->add_bottom("x_static");
+    x_static_transform_param->add_top("W_xc_x_static_preshape");
+    x_static_transform_param->add_propagate_down(true);
+
+    LayerParameter* reshape_param = net_param->add_layer();
+    reshape_param->set_type("Reshape");
+    BlobShape* new_shape =
+         reshape_param->mutable_reshape_param()->mutable_shape();
+    new_shape->add_dim(1);  // One timestep.
+    // Should infer this->N as the dimension so we can reshape on batch size.
+    new_shape->add_dim(-1);
+    new_shape->add_dim(
+        x_static_transform_param->inner_product_param().num_output());
+    reshape_param->set_name("W_xc_x_static_reshape");
+    reshape_param->add_bottom("W_xc_x_static_preshape");
+    reshape_param->add_top("W_xc_x_static");
+  }
+
+  LayerParameter* x_slice_param = net_param->add_layer();
+  x_slice_param->CopyFrom(slice_param);
+  x_slice_param->add_bottom("W_xc_x");
+  x_slice_param->set_name("W_xc_x_slice");
+
+  LayerParameter output_concat_layer;
+  output_concat_layer.set_name("h_concat");
+  output_concat_layer.set_type("Concat");
+  output_concat_layer.add_top("h");
+  output_concat_layer.mutable_concat_param()->set_axis(0);
+
+  for (int t = 1; t <= this->T_; ++t) {
+    string ts = format_int(t);
+    cont_slice_param->add_top("cont_" + ts);
+    x_slice_param->add_top("W_xc_x_" + ts);    
+  }
+
+  int depth = this->layer_param_.recurrent_param().depth();
+  if (depth==0) {
+      depth = this->T_;
+  }
+  for (int t = 1; t <= depth*this->T_; ++t) {
+
+    int m = (t-1) % this->T_ + 1;
+    string ms = format_int(m);
+    string ts  = format_int(t);  
+    string tm1s = format_int(t - 1);     
+
+    // Add layers to flush the hidden state when beginning a new
+    // sequence, as indicated by cont_t.
+    //     h_conted_{t-1} := cont_t * h_{t-1}
+    //
+    // Normally, cont_t is binary (i.e., 0 or 1), so:
+    //     h_conted_{t-1} := h_{t-1} if cont_t == 1
+    //                       0   otherwise
+    {
+      LayerParameter* cont_h_param = net_param->add_layer();
+      cont_h_param->CopyFrom(scale_param);
+      cont_h_param->set_name("h_conted_" + tm1s);
+      cont_h_param->add_bottom("h_" + tm1s);
+      if (m == 1 && t >this->T_) {
+        cont_h_param->add_bottom("cont_2");
+      } else {
+        cont_h_param->add_bottom("cont_" + ms);
+      }
+      cont_h_param->add_top("h_conted_" + tm1s);
+    }
+
+    // Add layer to compute
+    //     W_hc_h_{t-1} := W_hc * h_conted_{t-1}
+    {
+      LayerParameter* w_param = net_param->add_layer();
+      w_param->CopyFrom(hidden_param);
+      w_param->set_name("transform_" + ts);
+      w_param->add_param()->set_name("W_hc");
+      w_param->add_bottom("h_conted_" + tm1s);
+      w_param->add_top("W_hc_h_" + tm1s);
+      w_param->mutable_inner_product_param()->set_axis(2);
+    }
+
+    // Add the outputs of the linear transformations to compute the gate input.
+    //     gate_input_t := W_hc * h_conted_{t-1} + W_xc * x_t + b_c
+    //                   = W_hc_h_{t-1} + W_xc_x_t + b_c
+    {
+      LayerParameter* input_sum_layer = net_param->add_layer();
+      input_sum_layer->CopyFrom(sum_param);
+      input_sum_layer->set_name("gate_input_" + ts);
+      input_sum_layer->add_bottom("W_hc_h_" + tm1s);
+      input_sum_layer->add_bottom("W_xc_x_" + ms);
+      if (this->static_input_) {
+        input_sum_layer->add_bottom("W_xc_x_static");
+      }
+      input_sum_layer->add_top("gate_input_" + ts);
+    }
+
+    // Add LSTMUnit layer to compute the cell & hidden vectors c_t and h_t.
+    // Inputs: c_{t-1}, gate_input_t = (i_t, f_t, o_t, g_t), cont_t
+    // Outputs: c_t, h_t
+    //     [ i_t' ]
+    //     [ f_t' ] := gate_input_t
+    //     [ o_t' ]
+    //     [ g_t' ]
+    //         i_t := \sigmoid[i_t']
+    //         f_t := \sigmoid[f_t']
+    //         o_t := \sigmoid[o_t']
+    //         g_t := \tanh[g_t']
+    //         c_t := cont_t * (f_t .* c_{t-1}) + (i_t .* g_t)
+    //         h_t := o_t .* \tanh[c_t]
+    {
+      LayerParameter* lstm_unit_param = net_param->add_layer();
+      lstm_unit_param->set_type("LSTMUnit");
+      lstm_unit_param->add_bottom("c_" + tm1s);
+      lstm_unit_param->add_bottom("gate_input_" + ts);
+      if (m == 1 && t >this->T_) {
+        lstm_unit_param->add_bottom("cont_2");
+      } else {
+        lstm_unit_param->add_bottom("cont_" + ms);
+      }
+      lstm_unit_param->add_top("c_" + ts);
+      lstm_unit_param->add_top("h_" + ts);
+      lstm_unit_param->set_name("unit_" + ts);
+    }
+
+  }  
+  for (int t = 1; t <= this->T_; ++t)
+  {
+    output_concat_layer.add_bottom("h_" + format_int((depth-1)*this->T_+t));
+  } 
+
+  {
+    LayerParameter* c_T_copy_param = net_param->add_layer();
+    c_T_copy_param->CopyFrom(split_param);
+    c_T_copy_param->add_bottom("c_" + format_int(depth*this->T_));
+    c_T_copy_param->add_top("c_T");
+  }
+  net_param->add_layer()->CopyFrom(output_concat_layer);
+}
+
+INSTANTIATE_CLASS(CircleLSTMLayer);
+REGISTER_LAYER_CLASS(CircleLSTM);
+
+}  // namespace caffe
+