CVPR Experiments

Joseph Redmon

2015-11-04 8fd18add6e060a433629fae3fa2a7ef75df4644e

 src/parser.c

@@ -124,8 +124,9 @@
    c = params.c;
    batch=params.batch;
    if(!(h && w && c)) error("Layer before convolutional layer must output image.");
    int batch_normalize = option_find_int_quiet(options, "batch_normalize", 0);

    convolutional_layer layer = make_convolutional_layer(batch,h,w,c,n,size,stride,pad,activation);
    convolutional_layer layer = make_convolutional_layer(batch,h,w,c,n,size,stride,pad,activation, batch_normalize);

    char *weights = option_find_str(options, "weights", 0);
    char *biases = option_find_str(options, "biases", 0);
@@ -227,6 +228,7 @@
    int noadjust = option_find_int_quiet(options, "noadjust",0);

    crop_layer l = make_crop_layer(batch,h,w,c,crop_height,crop_width,flip, angle, saturation, exposure);
    l.shift = option_find_float(options, "shift", 0);
    l.noadjust = noadjust;
    return l;
}
@@ -452,6 +454,7 @@
            fprintf(stderr, "Type not recognized: %s\n", s->type);
        }
        l.dontload = option_find_int_quiet(options, "dontload", 0);
        l.dontloadscales = option_find_int_quiet(options, "dontloadscales", 0);
        option_unused(options);
        net.layers[count] = l;
        free_section(s);
@@ -633,19 +636,13 @@
#endif
            int num = l.n*l.c*l.size*l.size;
            fwrite(l.biases, sizeof(float), l.n, fp);
            fwrite(l.filters, sizeof(float), num, fp);
        }
        if(l.type == DECONVOLUTIONAL){
#ifdef GPU
            if(gpu_index >= 0){
                pull_deconvolutional_layer(l);
            if (l.batch_normalize){
                fwrite(l.scales, sizeof(float), l.n, fp);
                fwrite(l.rolling_mean, sizeof(float), l.n, fp);
                fwrite(l.rolling_variance, sizeof(float), l.n, fp);
            }
#endif
            int num = l.n*l.c*l.size*l.size;
            fwrite(l.biases, sizeof(float), l.n, fp);
            fwrite(l.filters, sizeof(float), num, fp);
        }
        if(l.type == CONNECTED){
        } if(l.type == CONNECTED){
#ifdef GPU
            if(gpu_index >= 0){
                pull_connected_layer(l);
@@ -682,6 +679,11 @@
        if(l.type == CONVOLUTIONAL){
            int num = l.n*l.c*l.size*l.size;
            fread(l.biases, sizeof(float), l.n, fp);
            if (l.batch_normalize && (!l.dontloadscales)){
                fread(l.scales, sizeof(float), l.n, fp);
                fread(l.rolling_mean, sizeof(float), l.n, fp);
                fread(l.rolling_variance, sizeof(float), l.n, fp);
            }
            fread(l.filters, sizeof(float), num, fp);
#ifdef GPU
            if(gpu_index >= 0){