stuff

src/activation_kernels.cu

@@ -7,6 +7,13 @@ extern "C" {
 #include "cuda.h"
 }
 
+
+__device__ float hardtan_activate_kernel(float x)
+{
+    if (x < -1) return -1;
+    if (x > 1) return 1;
+    return x;
+}
 __device__ float linear_activate_kernel(float x){return x;}
 __device__ float logistic_activate_kernel(float x){return 1./(1. + exp(-x));}
 __device__ float loggy_activate_kernel(float x){return 2./(1. + exp(-x)) - 1;}
@@ -29,6 +36,12 @@ __device__ float stair_activate_kernel(float x)
     else return (x - n) + floor(x/2.);
 }
  
+
+__device__ float hardtan_gradient_kernel(float x)
+{
+    if (x > -1 && x < 1) return 1;
+    return 0;
+}
 __device__ float linear_gradient_kernel(float x){return 1;}
 __device__ float logistic_gradient_kernel(float x){return (1-x)*x;}
 __device__ float loggy_gradient_kernel(float x)
@@ -74,6 +87,8 @@ __device__ float activate_kernel(float x, ACTIVATION a)
             return plse_activate_kernel(x);
         case STAIR:
             return stair_activate_kernel(x);
+        case HARDTAN:
+            return hardtan_activate_kernel(x);
     }
     return 0;
 }
@@ -103,6 +118,8 @@ __device__ float gradient_kernel(float x, ACTIVATION a)
             return plse_gradient_kernel(x);
         case STAIR:
             return stair_gradient_kernel(x);
+        case HARDTAN:
+            return hardtan_gradient_kernel(x);
     }
     return 0;
 }

src/activations.c

@@ -30,6 +30,8 @@ char *get_activation_string(ACTIVATION a)
             return "leaky";
         case STAIR:
             return "stair";
+        case HARDTAN:
+            return "hardtan";
         default:
             break;
     }
@@ -44,6 +46,7 @@ ACTIVATION get_activation(char *s)
     if (strcmp(s, "elu")==0) return ELU;
     if (strcmp(s, "relie")==0) return RELIE;
     if (strcmp(s, "plse")==0) return PLSE;
+    if (strcmp(s, "hardtan")==0) return HARDTAN;
     if (strcmp(s, "linear")==0) return LINEAR;
     if (strcmp(s, "ramp")==0) return RAMP;
     if (strcmp(s, "leaky")==0) return LEAKY;
@@ -78,6 +81,8 @@ float activate(float x, ACTIVATION a)
             return plse_activate(x);
         case STAIR:
             return stair_activate(x);
+        case HARDTAN:
+            return hardtan_activate(x);
     }
     return 0;
 }
@@ -115,6 +120,8 @@ float gradient(float x, ACTIVATION a)
             return plse_gradient(x);
         case STAIR:
             return stair_gradient(x);
+        case HARDTAN:
+            return hardtan_gradient(x);
     }
     return 0;
 }

src/activations.h

@@ -4,7 +4,7 @@
 #include "math.h"
 
 typedef enum{
-    LOGISTIC, RELU, RELIE, LINEAR, RAMP, TANH, PLSE, LEAKY, ELU, LOGGY, STAIR
+    LOGISTIC, RELU, RELIE, LINEAR, RAMP, TANH, PLSE, LEAKY, ELU, LOGGY, STAIR, HARDTAN
 }ACTIVATION;
 
 ACTIVATION get_activation(char *s);
@@ -25,6 +25,12 @@ static inline float stair_activate(float x)
     if (n%2 == 0) return floor(x/2.);
     else return (x - n) + floor(x/2.);
 }
+static inline float hardtan_activate(float x)
+{
+    if (x < -1) return -1;
+    if (x > 1) return 1;
+    return x;
+}
 static inline float linear_activate(float x){return x;}
 static inline float logistic_activate(float x){return 1./(1. + exp(-x));}
 static inline float loggy_activate(float x){return 2./(1. + exp(-x)) - 1;}
@@ -41,6 +47,11 @@ static inline float plse_activate(float x)
     return .125*x + .5;
 }
 
+static inline float hardtan_gradient(float x)
+{
+    if (x > -1 && x < 1) return 1;
+    return 0;
+}
 static inline float linear_gradient(float x){return 1;}
 static inline float logistic_gradient(float x){return (1-x)*x;}
 static inline float loggy_gradient(float x)

src/classifier.c

@@ -477,6 +477,7 @@ void predict_classifier(char *datacfg, char *cfgfile, char *weightfile, char *fi
     int *indexes = calloc(top, sizeof(int));
     char buff[256];
     char *input = buff;
+    int size = net.w;
     while(1){
         if(filename){
             strncpy(input, filename, 256);
@@ -487,8 +488,12 @@ void predict_classifier(char *datacfg, char *cfgfile, char *weightfile, char *fi
             if(!input) return;
             strtok(input, "\n");
         }
-        image im = load_image_color(input, net.w, net.h);
+        image im = load_image_color(input, 0, 0);
-        float *X = im.data;
+        image r = resize_min(im, size);
+        resize_network(&net, r.w, r.h);
+        printf("%d %d\n", r.w, r.h);
+
+        float *X = r.data;
         time=clock();
         float *predictions = network_predict(net, X);
         top_predictions(net, top, indexes);
@@ -497,6 +502,7 @@ void predict_classifier(char *datacfg, char *cfgfile, char *weightfile, char *fi
             int index = indexes[i];
             printf("%s: %f\n", names[index], predictions[index]);
         }
+        if(r.data != im.data) free_image(r);
         free_image(im);
         if (filename) break;
     }

src/convolutional_kernels.cu

@@ -142,6 +142,7 @@ void backward_convolutional_layer_gpu(convolutional_layer l, network_state state
     if(l.batch_normalize){
         backward_batchnorm_layer_gpu(l, state);
     }
+    float *original_input = state.input;
 
     if(l.xnor) state.input = l.binary_input_gpu;
 #ifdef CUDNN
@@ -176,6 +177,7 @@ void backward_convolutional_layer_gpu(convolutional_layer l, network_state state
                 l.dsrcTensorDesc,
                 state.delta);
         if(l.binary || l.xnor) swap_binary(&l);
+        if(l.xnor) gradient_array_ongpu(original_input, l.batch*l.c*l.h*l.w, HARDTAN, state.delta);
     }
 
 #else
@@ -197,7 +199,10 @@ void backward_convolutional_layer_gpu(convolutional_layer l, network_state state
             gemm_ongpu(1,0,n,k,m,1,a,n,b + i*k*m,k,0,c,k);
 
             col2im_ongpu(state.workspace, l.c,  l.h,  l.w,  l.size,  l.stride, l.pad, state.delta + i*l.c*l.h*l.w);
-            if(l.binary || l.xnor) swap_binary(&l);
+            if(l.binary || l.xnor) {
+                swap_binary(&l);
+            }
+            if(l.xnor) gradient_array_ongpu(original_input + i*l.c*l.h*l.w, l.c*l.h*l.w, HARDTAN, state.delta + i*l.c*l.h*l.w);
         }
     }
 #endif

src/network.c

@@ -434,7 +434,7 @@ int resize_network(network *net, int w, int h)
         net->workspace = cuda_make_array(0, (workspace_size-1)/sizeof(float)+1);
 #else
         free(net->workspace);
-        net->workspace = calloc(1, (workspace_size-1)/sizeof(float)+1);
+        net->workspace = calloc(1, workspace_size);
 #endif
     //fprintf(stderr, " Done!\n");
     return 0;

src/tag.c

@@ -125,7 +125,7 @@ void test_tag(char *cfgfile, char *weightfile, char *filename)
             int index = indexes[i];
             printf("%.1f%%: %s\n", predictions[index]*100, names[index]);
         }
-        free_image(r);
+        if(r.data != im.data) free_image(r);
         free_image(im);
         if (filename) break;
     }