1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
|
#include "Halide.h"
#include "halide_trace_config.h"
namespace {
class BilateralGrid : public Halide::Generator<BilateralGrid> {
public:
GeneratorParam<int> s_sigma{"s_sigma", 8};
Input<Buffer<float, 2>> input{"input"};
Input<float> r_sigma{"r_sigma"};
Output<Buffer<float, 2>> bilateral_grid{"bilateral_grid"};
void generate() {
Var x("x"), y("y"), z("z"), c("c");
// Add a boundary condition
Func clamped = Halide::BoundaryConditions::repeat_edge(input);
// Construct the bilateral grid
RDom r(0, s_sigma, 0, s_sigma);
Expr val = clamped(x * s_sigma + r.x - s_sigma / 2, y * s_sigma + r.y - s_sigma / 2);
val = clamp(val, 0.0f, 1.0f);
Expr zi = cast<int>(val * (1.0f / r_sigma) + 0.5f);
Func histogram("histogram");
histogram(x, y, z, c) = 0.0f;
histogram(x, y, zi, c) += mux(c, {val, 1.0f});
// Blur the grid using a five-tap filter
Func blurx("blurx"), blury("blury"), blurz("blurz");
blurz(x, y, z, c) = (histogram(x, y, z - 2, c) +
histogram(x, y, z - 1, c) * 4 +
histogram(x, y, z, c) * 6 +
histogram(x, y, z + 1, c) * 4 +
histogram(x, y, z + 2, c));
blurx(x, y, z, c) = (blurz(x - 2, y, z, c) +
blurz(x - 1, y, z, c) * 4 +
blurz(x, y, z, c) * 6 +
blurz(x + 1, y, z, c) * 4 +
blurz(x + 2, y, z, c));
blury(x, y, z, c) = (blurx(x, y - 2, z, c) +
blurx(x, y - 1, z, c) * 4 +
blurx(x, y, z, c) * 6 +
blurx(x, y + 1, z, c) * 4 +
blurx(x, y + 2, z, c));
// Take trilinear samples to compute the output
val = clamp(input(x, y), 0.0f, 1.0f);
Expr zv = val * (1.0f / r_sigma);
zi = cast<int>(zv);
Expr zf = zv - zi;
Expr xf = cast<float>(x % s_sigma) / s_sigma;
Expr yf = cast<float>(y % s_sigma) / s_sigma;
Expr xi = x / s_sigma;
Expr yi = y / s_sigma;
Func interpolated("interpolated");
interpolated(x, y, c) =
lerp(lerp(lerp(blury(xi, yi, zi, c), blury(xi + 1, yi, zi, c), xf),
lerp(blury(xi, yi + 1, zi, c), blury(xi + 1, yi + 1, zi, c), xf), yf),
lerp(lerp(blury(xi, yi, zi + 1, c), blury(xi + 1, yi, zi + 1, c), xf),
lerp(blury(xi, yi + 1, zi + 1, c), blury(xi + 1, yi + 1, zi + 1, c), xf), yf),
zf);
// Normalize
bilateral_grid(x, y) = interpolated(x, y, 0) / interpolated(x, y, 1);
/* ESTIMATES */
// (This can be useful in conjunction with RunGen and benchmarks as well
// as auto-schedule, so we do it in all cases.)
// Provide estimates on the input image
input.set_estimates({{0, 1536}, {0, 2560}});
// Provide estimates on the parameters
r_sigma.set_estimate(0.1f);
// TODO: Compute estimates from the parameter values
histogram.set_estimate(z, -2, 16);
blurz.set_estimate(z, 0, 12);
blurx.set_estimate(z, 0, 12);
blury.set_estimate(z, 0, 12);
bilateral_grid.set_estimates({{0, 1536}, {0, 2560}});
if (using_autoscheduler()) {
// nothing
} else if (get_target().has_gpu_feature()) {
// 0.50ms on an RTX 2060
Var xi("xi"), yi("yi"), zi("zi");
// Schedule blurz in 8x8 tiles. This is a tile in
// grid-space, which means it represents something like
// 64x64 pixels in the input (if s_sigma is 8).
blurz.compute_root().reorder(c, z, x, y).gpu_tile(x, y, xi, yi, 8, 8);
// Schedule histogram to happen per-tile of blurz, with
// intermediate results in shared memory. This means histogram
// and blurz makes a three-stage kernel:
// 1) Zero out the 8x8 set of histograms
// 2) Compute those histogram by iterating over lots of the input image
// 3) Blur the set of histograms in z
histogram.reorder(c, z, x, y).compute_at(blurz, x).gpu_threads(x, y);
histogram.update().reorder(c, r.x, r.y, x, y).gpu_threads(x, y).unroll(c);
// Schedule the remaining blurs and the sampling at the end similarly.
blurx
.compute_root()
.reorder(c, x, y, z)
.reorder_storage(c, x, y, z)
.vectorize(c)
.unroll(y, 2, TailStrategy::RoundUp)
.gpu_tile(x, y, z, xi, yi, zi, 32, 8, 1, TailStrategy::RoundUp);
blury
.compute_root()
.reorder(c, x, y, z)
.reorder_storage(c, x, y, z)
.vectorize(c)
.unroll(y, 2, TailStrategy::RoundUp)
.gpu_tile(x, y, z, xi, yi, zi, 32, 8, 1, TailStrategy::RoundUp);
bilateral_grid.compute_root().gpu_tile(x, y, xi, yi, 32, 8);
interpolated.compute_at(bilateral_grid, xi).vectorize(c);
} else {
// CPU schedule.
// 2.04ms on an Intel i9-9960X using 32 threads at 3.5 GHz using
// target "host". This is a little less SIMD-friendly than some of the
// other apps, so we benefit from hyperthreading, and don't benefit
// from 512-bit vectors.
blurz.compute_at(blurx, y)
.reorder(c, z, x, y)
.vectorize(x, 8)
.unroll(c);
histogram.compute_at(blurz, y);
histogram.update()
.reorder(c, r.x, r.y, x, y)
.unroll(c);
blurx.compute_root()
.reorder(c, x, z, y)
.parallel(y)
.vectorize(x, 8)
.unroll(c);
blury.compute_at(bilateral_grid, y)
.store_in(MemoryType::Stack)
.reorder(c, x, y, z)
.reorder_storage(c, z, x, y)
.vectorize(x, 8)
.unroll(c);
bilateral_grid.compute_root()
.parallel(y, 8)
.vectorize(x, 8);
}
/* Optional tags to specify layout for HalideTraceViz */
{
Halide::Trace::FuncConfig cfg;
cfg.pos.x = 100;
cfg.pos.y = 300;
input.add_trace_tag(cfg.to_trace_tag());
cfg.pos.x = 1564;
bilateral_grid.add_trace_tag(cfg.to_trace_tag());
}
{
Halide::Trace::FuncConfig cfg;
cfg.strides = {{1, 0}, {0, 1}, {40, 0}};
cfg.zoom = 3;
cfg.max = 32;
cfg.pos.x = 550;
cfg.pos.y = 100;
histogram.add_trace_tag(cfg.to_trace_tag());
cfg.max = 512;
cfg.pos.y = 300;
blurz.add_trace_tag(cfg.to_trace_tag());
cfg.max = 8192;
cfg.pos.y = 500;
blurx.add_trace_tag(cfg.to_trace_tag());
cfg.max = 131072;
cfg.pos.y = 700;
blury.add_trace_tag(cfg.to_trace_tag());
}
{
// GlobalConfig applies to the entire visualization pipeline;
// you can set this tag on any Func that is realized, but only
// the last one seen will be used. (Since the tags are emitted in
// an arbitrary order, emitting only one such tag is the best practice).
// Note also that since the global settings are often context-dependent
// (eg the output size and timestep may vary depending on the
// input data), it's often more useful to specify these on the
// command line.
Halide::Trace::GlobalConfig global_cfg;
global_cfg.timestep = 1000;
bilateral_grid.add_trace_tag(global_cfg.to_trace_tag());
}
}
};
} // namespace
HALIDE_REGISTER_GENERATOR(BilateralGrid, bilateral_grid)
|
