Skip to content

Commit

Permalink
Merge output matching expect for 5 leftmost columns
Browse files Browse the repository at this point in the history
  • Loading branch information
@mcoduoza
mcoduoza committed May 30, 2024
1 parent d119ffd commit b0d38da
Show file tree
Hide file tree
Showing 4 changed files with 57 additions and 20 deletions.
5 changes: 3 additions & 2 deletions apps/hardware_benchmarks/apps/hdr_plus/compare_divided_tiles.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -73,11 +73,12 @@

abs_difference = abs(value - kayvon_divided_tiles[ty][tx][yi][xi])
if(abs_difference > tolerance):
print(f"At tx {tx}, ty {tx}, xi {xi}, yi {yi}, expected {kayvon_divided_tiles[ty][tx][yi][xi]} but got {value}. Difference is {abs_difference}.")
print(f"At tx {tx}, ty {ty}, xi {xi}, yi {yi}, expected {kayvon_divided_tiles[ty][tx][yi][xi]} but got {value}. Difference is {abs_difference}.")

line = trace_file.readline()
lineCount += 1
if(lineCount == 256):
#if(lineCount == 204800):
if(lineCount == 3215099):
break

#print(align_coords)
Expand Down
10 changes: 6 additions & 4 deletions apps/hardware_benchmarks/apps/hdr_plus/compare_weighted_tiles.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -47,7 +47,8 @@

k_weighted_tiles_file.close()

#print(kayvon_weighted_tiles)
#print(kayvon_weighted_tiles[12][12][0][0])



# Now, compare to my weighted tiles
Expand All @@ -73,12 +74,13 @@

abs_difference = abs(value - kayvon_weighted_tiles[ty][tx][yi][xi])
if(abs_difference > tolerance):
print(f"At tx {tx}, ty {tx}, xi {xi}, yi {yi}, expected {kayvon_weighted_tiles[ty][tx][yi][xi]} but got {value}. Difference is {abs_difference}.")
print(f"At tx {tx}, ty {ty}, xi {xi}, yi {yi}, expected {kayvon_weighted_tiles[ty][tx][yi][xi]} but got {value}. Difference is {abs_difference}.")

line = trace_file.readline()
lineCount += 1
if(lineCount == 20480):
break
#if(lineCount == 256):
#if(lineCount == 4343536):
# break

#print(align_coords)
trace_file.close()
Expand Down
2 changes: 1 addition & 1 deletion apps/hardware_benchmarks/apps/hdr_plus/compare_weights.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@



trace_file_path = "./weights_out.txt"
trace_file_path = "./out.txt"


num_tx = 156
Expand Down
60 changes: 47 additions & 13 deletions apps/hardware_benchmarks/apps/hdr_plus/hdr_plus_generator.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -1031,20 +1031,34 @@ class HDRPlus : public Halide::Generator<HDRPlus> {
// alt_val = deinterleaved(alt_x, alt_y, c, n);


ref_x = clamp(tx_image*(T_SIZE/2) + xi, 0, input.width() - 1);
ref_y = clamp(ty_image*(T_SIZE/2) + yi, 0, input.height() - 1);
alt_x = clamp(tx_image*(T_SIZE/2) + xi + 2*offset_x, 0, input.width() - 1);
alt_y = clamp(ty_image*(T_SIZE/2) + yi+ 2*offset_y, 0, input.height() - 1);
ref_val = clamped_input_float(ref_x, ref_y, 0);
alt_val = clamped_input_float(alt_x, alt_y, n);
// ref_x = clamp((tx_image*(T_SIZE/2)) + xi, 0, input.width() - 1);
// ref_y = clamp((ty_image*(T_SIZE/2)) + yi, 0, input.height() - 1);
// alt_x = clamp((tx_image*(T_SIZE/2)) + xi + (2*offset_x), 0, input.width() - 1);
// alt_y = clamp((ty_image*(T_SIZE/2)) + yi+ (2*offset_y), 0, input.height() - 1);
ref_x = (tx_image*(T_SIZE/2)) + xi;
ref_y = (ty_image*(T_SIZE/2)) + yi;
alt_x = (tx_image*(T_SIZE/2)) + xi + (2*offset_x);
alt_y = (ty_image*(T_SIZE/2)) + yi+ (2*offset_y);
//ref_val = clamped_input_float(ref_x, ref_y, 0);
//alt_val = clamped_input_float(alt_x, alt_y, n);
ref_val = clamped_input(ref_x, ref_y, 0);
alt_val = clamped_input(alt_x, alt_y, n);

Expr x_index = select(n == 0, ref_x, alt_x);
Expr y_index = select(n == 0, ref_y, alt_y);





Func val;
//val(xi, yi, tx, ty, c, n) = select(n == 0, ref_val, alt_val);
val(xi, yi, tx_image, ty_image, n) = select(n == 0, ref_val, alt_val);
//val(xi, yi, tx_image, ty_image, n) = select(n == 0, ref_val, alt_val);
//val(xi, yi, tx_image, ty_image, n) = clamped_input(x_index, y_index, n);
Expr x_index_in_bounds = ((x_index >= 0) && (x_index < input.width()));
Expr y_index_in_bounds = ((y_index >= 0) && (y_index < input.height()));

val(xi, yi, tx_image, ty_image, n) = select(x_index_in_bounds && y_index_in_bounds, clamped_input(x_index, y_index, n), 0.0f);
//val.trace_stores();
// Weighted sum of all frames (reference frame and all alternate frames)
// TODO: Unshuffle back into bayer pattern before sending output
Expand All @@ -1070,7 +1084,13 @@ class HDRPlus : public Halide::Generator<HDRPlus> {
//output_deinterleaved_tiled(xi, yi, tx, ty, c) = u16(sum(u32(unscaled_normalized_weight(tx, ty, c, r_imgs) * val(xi, yi, tx, ty, c, r_imgs))));
//output_deinterleaved_tiled(x, y, c) = sum(unscaled_normalized_weight(tile_x, tile_y, c, r_imgs) * val(ix, iy, tile_x, tile_y, c, r_imgs));
output_tiled(xi, yi, tx_image, ty_image) = sum(weight(tx_image, ty_image, r_imgs) * val(xi, yi, tx_image, ty_image, r_imgs));
output_tiled.trace_stores();
//output_tiled(xi, yi, tx_image, ty_image) = cast<float>(0);

//output_tiled(xi, yi, tx_image, ty_image) += select((x_index_in_bounds && y_index_in_bounds), weight(tx_image, ty_image, n) * val(xi, yi, tx_image, ty_image, n), 0.0f);
// output_tiled(xi, yi, tx_image, ty_image) = weight(tx_image, ty_image, 0) * val(xi, yi, tx_image, ty_image, 0)
// + weight(tx_image, ty_image, 1) * val(xi, yi, tx_image, ty_image, 1)
// + weight(tx_image, ty_image, 2) * val(xi, yi, tx_image, ty_image, 2);
//output_tiled.trace_stores();
//output_deinterleaved_tiled.trace_stores();

//output_deinterleaved_tiled(xi, yi, tx, ty, c) = u16(sum(u32(mul2(unscaled_normalized_weight(tx, ty, c, r_imgs), val(xi, yi, tx, ty, c, r_imgs)))));
Expand All @@ -1082,7 +1102,7 @@ class HDRPlus : public Halide::Generator<HDRPlus> {


Func output_tiled_normalized_cosined;
output_tiled_normalized_cosined(xi, yi, tx_image, ty_image) = output_tiled(xi, yi, tx_image, ty_image) * raised_cosine_weight(xi) * raised_cosine_weight(yi) * 1.0f/sum_weight(tx_image, ty_image);
output_tiled_normalized_cosined(xi, yi, tx_image, ty_image) = output_tiled(xi, yi, tx_image, ty_image) * raised_cosine_weight(xi) * raised_cosine_weight(yi) * (1.0f/sum_weight(tx_image, ty_image));
//output_tiled_normalized_cosined.trace_stores();


Expand All @@ -1099,10 +1119,22 @@ class HDRPlus : public Halide::Generator<HDRPlus> {
Expr num_ty_locs = 139;

RDom tile_RDom(0, num_tx_locs, 0, num_ty_locs, 0, T_SIZE, 0, T_SIZE);
Expr x_prime = tile_RDom.x * T_SIZE/2 + tile_RDom.z;
Expr y_prime = tile_RDom.y * T_SIZE/2 + tile_RDom.w;
Expr x_prime = (tile_RDom.x * (T_SIZE/2)) + tile_RDom.z;
Expr y_prime = (tile_RDom.y * (T_SIZE/2)) + tile_RDom.w;
//final_output(x_prime, y_prime, c) += mul2(final_output_tiled(tile_RDom.z, tile_RDom.w, tile_RDom.x, tile_RDom.y, c), mul2(cast<uint16_t>(raised_cosine_weight(tile_RDom.z) * 256.0f), cast<uint16_t>(raised_cosine_weight(tile_RDom.w) * 256.0f)));
final_output(x_prime, y_prime) += output_tiled_normalized_cosined(tile_RDom.z, tile_RDom.w, tile_RDom.x, tile_RDom.y);


Expr track_value = ((x_prime == 0) && (y_prime == 129));
Expr my_debug_value = output_tiled_normalized_cosined(tile_RDom.z, tile_RDom.w, tile_RDom.x, tile_RDom.y);
my_debug_value = print_when((x_prime == 0) && (y_prime == 129), my_debug_value, "This is OTNC when x = 0, y = 129");

//Func my_debug_tile;
//my_debug_tile(x, y) = cast<float>(0);


//my_debug_tile(x_prime, y_prime) = select(track_value, output_tiled_normalized_cosined(tile_RDom.z, tile_RDom.w, tile_RDom.x, tile_RDom.y), 0.0f);
//my_debug_tile.trace_stores();



Expand All @@ -1129,6 +1161,7 @@ class HDRPlus : public Halide::Generator<HDRPlus> {
Func merge_output;
//merge_output(x, y) = u16(select((y%2)==0, row_r_result(x, y/2), row_b_result(x, y/2)));
merge_output(x, y) = final_output(x, y);
merge_output.trace_stores();
output(x, y, c) = u8(merge_output(x, y) * 255.f);
//output.trace_stores();
//output(x, y) = input(x, y, 0);
Expand Down Expand Up @@ -1330,7 +1363,7 @@ class HDRPlus : public Halide::Generator<HDRPlus> {

//MERGE SCHEDULE
//output_shuffle.reorder(c, x, y).tile(x, y, xo, yo, xi, yi, 64, 64).fuse(xo, yo, outer).parallel(outer);
merge_output.reorder(x, y).tile(x, y, xo, yo, xi, yi, 16, 16).fuse(xo, yo, outer).parallel(outer);
merge_output.reorder(x, y).tile(x, y, xo, yo, xi, yi, 64, 64).fuse(xo, yo, outer).parallel(outer);
merge_output.compute_root();
//dist_channel.reorder(c, tx, ty, n).compute_root();
dist_channel.reorder(tx_image, ty_image, n).compute_root();
Expand All @@ -1343,6 +1376,7 @@ class HDRPlus : public Halide::Generator<HDRPlus> {
output_tiled.store_at(merge_output, outer).compute_at(merge_output, yi);
output_tiled_normalized_cosined.store_at(merge_output, outer).compute_at(merge_output, yi);
final_output.store_at(merge_output, outer).compute_at(merge_output, yi);
//my_debug_tile.store_at(merge_output, outer).compute_at(merge_output, yi);


// // CAMERA PIPELINE SCHEDULE
Expand Down Expand Up @@ -1816,7 +1850,7 @@ class HDRPlus : public Halide::Generator<HDRPlus> {

Expr raised_cosine_weight(Expr in){
const float PI = 3.141592f;
return 0.5f - 0.5f * cos(2.0f * PI * (in + 0.5f) / T_SIZE);
return 0.5f - (0.5f * cos(2.0f * PI * (in + 0.5f) / T_SIZE));
}

/*
Expand Down

0 comments on commit b0d38da

Please sign in to comment.