Skip to content

Commit

Permalink
added some sse instructions for scd
Browse files Browse the repository at this point in the history
  • Loading branch information
@liuliu
liuliu committed Dec 13, 2014
1 parent 06c2fbf commit c233bf6
Show file tree
Hide file tree
Showing 2 changed files with 137 additions and 8 deletions.
6 changes: 6 additions & 0 deletions lib/ccv.h
View file
Original file line number Diff line number Diff line change
Expand Up @@ -26,6 +26,12 @@
#define ccrealloc realloc
#define ccfree free

#if _POSIX_C_SOURCE >= 200112L || _XOPEN_SOURCE >= 600
#define ccmemalign posix_memalign
#else
#define ccmemalign(memptr, alignment, size) (*memptr = memalign(alignment, size))
#endif

enum {
CCV_8U = 0x01000,
CCV_32S = 0x02000,
Expand Down
139 changes: 131 additions & 8 deletions lib/ccv_scd.c
View file
Original file line number Diff line number Diff line change
@@ -1,5 +1,10 @@
#include "ccv.h"
#include "ccv_internal.h"
#if defined(HAVE_SSE2)
#include <xmmintrin.h>
#elif defined(HAVE_NEON)
#include <arm_neon.h>
#endif
#ifdef HAVE_GSL
#include <gsl/gsl_rng.h>
#include <gsl/gsl_randist.h>
Expand Down Expand Up @@ -129,6 +134,66 @@ void ccv_scd(ccv_dense_matrix_t* a, ccv_dense_matrix_t** b, int type)
ccv_matrix_free(dv);
}

#if defined(HAVE_SSE2)
static inline void _ccv_scd_run_feature_at_sse2(float* at, int cols, ccv_scd_stump_feature_t* feature, __m128 surf[8])
{
int i;
// extract feature
#pragma unroll
for (i = 0; i < 4; i++)
{
__m128 d0 = _mm_loadu_ps(at + (cols * feature->sy[i] + feature->sx[i]) * 8);
__m128 d1 = _mm_loadu_ps(at + 4 + (cols * feature->sy[i] + feature->sx[i]) * 8);
__m128 du0 = _mm_loadu_ps(at + (cols * feature->dy[i] + feature->sx[i]) * 8);
__m128 du1 = _mm_loadu_ps(at + 4 + (cols * feature->dy[i] + feature->sx[i]) * 8);
__m128 dv0 = _mm_loadu_ps(at + (cols * feature->sy[i] + feature->dx[i]) * 8);
__m128 dv1 = _mm_loadu_ps(at + 4 + (cols * feature->sy[i] + feature->dx[i]) * 8);
__m128 duv0 = _mm_loadu_ps(at + (cols * feature->dy[i] + feature->dx[i]) * 8);
__m128 duv1 = _mm_loadu_ps(at + 4 + (cols * feature->dy[i] + feature->dx[i]) * 8);
surf[i * 2] = _mm_sub_ps(_mm_add_ps(duv0, d0), _mm_add_ps(du0, dv0));
surf[i * 2 + 1] = _mm_sub_ps(_mm_add_ps(duv1, d1), _mm_add_ps(du1, dv1));
}
// L2Hys normalization
__m128 v0 = _mm_add_ps(_mm_mul_ps(surf[0], surf[0]), _mm_mul_ps(surf[1], surf[1]));
__m128 v1 = _mm_add_ps(_mm_mul_ps(surf[2], surf[2]), _mm_mul_ps(surf[3], surf[3]));
__m128 v2 = _mm_add_ps(_mm_mul_ps(surf[4], surf[4]), _mm_mul_ps(surf[5], surf[5]));
__m128 v3 = _mm_add_ps(_mm_mul_ps(surf[6], surf[6]), _mm_mul_ps(surf[7], surf[7]));
v0 = _mm_add_ps(v0, v1);
v2 = _mm_add_ps(v2, v3);
union {
float f[4];
__m128 p;
} vx;
vx.p = _mm_add_ps(v0, v2);
v0 = _mm_set1_ps(1.0 / (sqrtf(vx.f[0] + vx.f[1] + vx.f[2] + vx.f[3]) + 1e-6));
static float thlf = -2.0 / 5.65685424949; // -sqrtf(32)
static float thuf = 2.0 / 5.65685424949; // sqrtf(32)
const __m128 thl = _mm_set1_ps(thlf);
const __m128 thu = _mm_set1_ps(thuf);
#pragma unroll
for (i = 0; i < 8; i++)
{
surf[i] = _mm_mul_ps(surf[i], v0);
surf[i] = _mm_min_ps(surf[i], thu);
surf[i] = _mm_max_ps(surf[i], thl);
}
__m128 u0 = _mm_add_ps(_mm_mul_ps(surf[0], surf[0]), _mm_mul_ps(surf[1], surf[1]));
__m128 u1 = _mm_add_ps(_mm_mul_ps(surf[2], surf[2]), _mm_mul_ps(surf[3], surf[3]));
__m128 u2 = _mm_add_ps(_mm_mul_ps(surf[4], surf[4]), _mm_mul_ps(surf[5], surf[5]));
__m128 u3 = _mm_add_ps(_mm_mul_ps(surf[6], surf[6]), _mm_mul_ps(surf[7], surf[7]));
u0 = _mm_add_ps(u0, u1);
u2 = _mm_add_ps(u2, u3);
union {
float f[4];
__m128 p;
} ux;
ux.p = _mm_add_ps(u0, u2);
u0 = _mm_set1_ps(1.0 / (sqrtf(ux.f[0] + ux.f[1] + ux.f[2] + ux.f[3]) + 1e-6));
#pragma unroll
for (i = 0; i < 8; i++)
surf[i] = _mm_mul_ps(surf[i], u0);
}
#else
static inline void _ccv_scd_run_feature_at(float* at, int cols, ccv_scd_stump_feature_t* feature, float surf[32])
{
int i, j;
Expand Down Expand Up @@ -159,6 +224,7 @@ static inline void _ccv_scd_run_feature_at(float* at, int cols, ccv_scd_stump_fe
for (i = 0; i < 32; i++)
surf[i] = surf[i] * u;
}
#endif

#ifdef HAVE_GSL
static ccv_array_t* _ccv_scd_collect_negatives(gsl_rng* rng, ccv_size_t size, ccv_array_t* hard_mine, int total, int grayscale)
Expand Down Expand Up @@ -339,7 +405,11 @@ static void _ccv_scd_precompute_feature_vectors(const ccv_array_t* features, con
{
ccv_scd_stump_feature_t* feature = (ccv_scd_stump_feature_t*)ccv_array_get(features, j);
// save to fv
#if defined(HAVE_SSE2)
_ccv_scd_run_feature_at_sse2(sat->data.f32, sat->cols, feature, (__m128*)_ccv_scd_get_surf_at(fv, j, i, positives->rnum, negatives->rnum));
#else
_ccv_scd_run_feature_at(sat->data.f32, sat->cols, feature, _ccv_scd_get_surf_at(fv, j, i, positives->rnum, negatives->rnum));
#endif
}
ccv_matrix_free(sat);
} parallel_endfor
Expand All @@ -358,7 +428,11 @@ static void _ccv_scd_precompute_feature_vectors(const ccv_array_t* features, con
{
ccv_scd_stump_feature_t* feature = (ccv_scd_stump_feature_t*)ccv_array_get(features, j);
// save to fv
#if defined(HAVE_SSE2)
_ccv_scd_run_feature_at_sse2(sat->data.f32, sat->cols, feature, (__m128*)_ccv_scd_get_surf_at(fv, j, i + positives->rnum, positives->rnum, negatives->rnum));
#else
_ccv_scd_run_feature_at(sat->data.f32, sat->cols, feature, _ccv_scd_get_surf_at(fv, j, i + positives->rnum, positives->rnum, negatives->rnum));
#endif
}
ccv_matrix_free(sat);
} parallel_endfor
Expand Down Expand Up @@ -605,7 +679,11 @@ static float _ccv_scd_threshold_at_hit_rate(double* s, int posnum, int negnum, f

static int _ccv_scd_classifier_cascade_pass(ccv_scd_classifier_cascade_t* cascade, ccv_dense_matrix_t* a)
{
#if defined(HAVE_SSE2)
__m128 surf[8];
#else
float surf[32];
#endif
ccv_dense_matrix_t* b = 0;
ccv_scd(a, &b, 0);
ccv_dense_matrix_t* sat = 0;
Expand All @@ -620,11 +698,28 @@ static int _ccv_scd_classifier_cascade_pass(ccv_scd_classifier_cascade_t* cascad
for (j = 0; j < classifier->count; j++)
{
ccv_scd_stump_feature_t* feature = classifier->features + j;
#if defined(HAVE_SSE2)
_ccv_scd_run_feature_at_sse2(sat->data.f32, sat->cols, feature, surf);
__m128 u0 = _mm_load1_ps(&feature->bias);
__m128 u1 = u0;
for (k = 0; k < 8; k += 2)
{
u0 = _mm_add_ps(u0, _mm_mul_ps(surf[k], _mm_loadu_ps(feature->w + k * 4)));
u1 = _mm_add_ps(u1, _mm_mul_ps(surf[k + 1], _mm_loadu_ps(feature->w + k * 4 + 4)));
}
struct {
float f[4];
__m128 packed;
} ux;
ux.packed = _mm_add_ps(u0, u1);
float u = expf(ux.f[0] + ux.f[1] + ux.f[2] + ux.f[3]);
#else
_ccv_scd_run_feature_at(sat->data.f32, sat->cols, feature, surf);
float u = feature->bias;
for (k = 0; k < 32; k++)
u += surf[k] * feature->w[k];
u = expf(u);
#endif
v += (u - 1) / (u + 1);
}
if (v <= classifier->threshold)
Expand Down Expand Up @@ -876,7 +971,7 @@ ccv_scd_classifier_cascade_t* ccv_scd_classifier_cascade_new(ccv_array_t* posfil
assert(z.pw);
z.nw = (double*)ccmalloc(sizeof(double) * negative_count);
assert(z.nw);
z.fv = (float*)ccmalloc(sizeof(float) * (z.positives->rnum + negative_count) * z.features->rnum * 32);
ccmemalign((void**)&z.fv, 16, sizeof(float) * (z.positives->rnum + negative_count) * z.features->rnum * 32);
assert(z.fv);
z.params = params;
ccv_function_state_begin(_ccv_scd_classifier_cascade_new_function_state_read, z, filename);
Expand Down Expand Up @@ -1253,7 +1348,7 @@ static int _ccv_is_equal_same_class(const void* _r1, const void* _r2, void* data

ccv_array_t* ccv_scd_detect_objects(ccv_dense_matrix_t* a, ccv_scd_classifier_cascade_t** cascades, int count, ccv_scd_param_t params)
{
int i, j, k, x, y, p, q, r;
int i, j, k, x, y, p, q;
int scale_upto = 1;
float up_ratio = 1.0;
for (i = 0; i < count; i++)
Expand All @@ -1273,7 +1368,11 @@ ccv_array_t* ccv_scd_detect_objects(ccv_dense_matrix_t* a, ccv_scd_classifier_ca
pyr[i] = 0;
ccv_sample_down(pyr[i - 1], &pyr[i], 0, 0, 0);
}
#if defined(HAVE_SSE2)
__m128 surf[8];
#else
float surf[32];
#endif
ccv_array_t** seq = (ccv_array_t**)alloca(sizeof(ccv_array_t*) * count);
for (i = 0; i < count; i++)
seq[i] = ccv_array_new(sizeof(ccv_comp_t), 64, 0);
Expand All @@ -1294,13 +1393,20 @@ ccv_array_t* ccv_scd_detect_objects(ccv_dense_matrix_t* a, ccv_scd_classifier_ca
ccv_dense_matrix_t* image = k == 0 ? pyr[i] : 0;
if (k > 0)
ccv_resample(pyr[i], &image, 0, rows, cols, CCV_INTER_AREA);
ccv_dense_matrix_t* bordered = 0;
ccv_border(image, (ccv_matrix_t**)&bordered, 0, cascade->margin);
if (k > 0)
ccv_matrix_free(image);
ccv_dense_matrix_t* scd = 0;
ccv_scd(bordered, &scd, 0);
ccv_matrix_free(bordered);
if (cascade->margin.left == 0 && cascade->margin.top == 0 && cascade->margin.right == 0 && cascade->margin.bottom == 0)
{
ccv_scd(image, &scd, 0);
if (k > 0)
ccv_matrix_free(image);
} else {
ccv_dense_matrix_t* bordered = 0;
ccv_border(image, (ccv_matrix_t**)&bordered, 0, cascade->margin);
if (k > 0)
ccv_matrix_free(image);
ccv_scd(bordered, &scd, 0);
ccv_matrix_free(bordered);
}
ccv_dense_matrix_t* sat = 0;
ccv_sat(scd, &sat, 0, CCV_PADDING_ZERO);
assert(CCV_GET_CHANNEL(sat->type) == 8);
Expand All @@ -1323,11 +1429,28 @@ ccv_array_t* ccv_scd_detect_objects(ccv_dense_matrix_t* a, ccv_scd_classifier_ca
for (q = 0; q < classifier->count; q++)
{
ccv_scd_stump_feature_t* feature = classifier->features + q;
#if defined(HAVE_SSE2)
_ccv_scd_run_feature_at_sse2(ptr + x * 8, sat->cols, feature, surf);
__m128 u0 = _mm_add_ps(_mm_mul_ps(surf[0], _mm_loadu_ps(feature->w)), _mm_mul_ps(surf[1], _mm_loadu_ps(feature->w + 4)));
__m128 u1 = _mm_add_ps(_mm_mul_ps(surf[2], _mm_loadu_ps(feature->w + 8)), _mm_mul_ps(surf[3], _mm_loadu_ps(feature->w + 12)));
__m128 u2 = _mm_add_ps(_mm_mul_ps(surf[4], _mm_loadu_ps(feature->w + 16)), _mm_mul_ps(surf[5], _mm_loadu_ps(feature->w + 20)));
__m128 u3 = _mm_add_ps(_mm_mul_ps(surf[6], _mm_loadu_ps(feature->w + 24)), _mm_mul_ps(surf[7], _mm_loadu_ps(feature->w + 28)));
u0 = _mm_add_ps(u0, u1);
u2 = _mm_add_ps(u2, u3);
union {
float f[4];
__m128 p;
} ux;
ux.p = _mm_add_ps(u0, u2);
float u = expf(feature->bias + ux.f[0] + ux.f[1] + ux.f[2] + ux.f[3]);
#else
_ccv_scd_run_feature_at(ptr + x * 8, sat->cols, feature, surf);
float u = feature->bias;
int r;
for (r = 0; r < 32; r++)
u += surf[r] * feature->w[r];
u = expf(u);
#endif
v += (u - 1) / (u + 1);
}
if (v <= classifier->threshold)
Expand Down

0 comments on commit c233bf6

Please sign in to comment.