Support Seg C++

pull/8/head
triple-Mu 2 years ago
parent 991f0885ab
commit 34ccc23a0e
  1. 55
      csrc/segment/CMakeLists.txt
  2. 107
      csrc/segment/include/config.h
  3. 133
      csrc/segment/include/utils.h
  4. 329
      csrc/segment/include/yolov8-seg.hpp
  5. 86
      csrc/segment/main.cpp
  6. 13
      export_seg.py
  7. 12
      infer.py
  8. 5
      models/common.py

@ -0,0 +1,55 @@
cmake_minimum_required(VERSION 2.8.12)
set(CMAKE_CUDA_ARCHITECTURES 60 61 62 70 72 75 86)
set(CMAKE_CUDA_COMPILER /usr/local/cuda/bin/nvcc)
project(yolov8-seg LANGUAGES CXX CUDA)
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++14 -O3 -g")
set(CMAKE_CXX_STANDARD 14)
set(CMAKE_BUILD_TYPE Debug)
option(CUDA_USE_STATIC_CUDA_RUNTIME OFF)
# CUDA
find_package(CUDA REQUIRED)
message(STATUS "CUDA Libs: \n${CUDA_LIBRARIES}\n")
message(STATUS "CUDA Headers: \n${CUDA_INCLUDE_DIRS}\n")
# OpenCV
find_package(OpenCV REQUIRED)
message(STATUS "OpenCV Libs: \n${OpenCV_LIBS}\n")
message(STATUS "OpenCV Libraries: \n${OpenCV_LIBRARIES}\n")
message(STATUS "OpenCV Headers: \n${OpenCV_INCLUDE_DIRS}\n")
# TensorRT
set(TensorRT_INCLUDE_DIRS /usr/include/x86_64-linux-gnu)
set(TensorRT_LIBRARIES /usr/lib/x86_64-linux-gnu)
message(STATUS "TensorRT Libs: \n${TensorRT_LIBRARIES}\n")
message(STATUS "TensorRT Headers: \n${TensorRT_INCLUDE_DIRS}\n")
list(APPEND INCLUDE_DIRS
${CUDA_INCLUDE_DIRS}
${OpenCV_INCLUDE_DIRS}
${TensorRT_INCLUDE_DIRS}
./include
)
list(APPEND ALL_LIBS
${CUDA_LIBRARIES}
${OpenCV_LIBRARIES}
${TensorRT_LIBRARIES}
)
include_directories(${INCLUDE_DIRS})
add_executable(${PROJECT_NAME}
main.cpp
include/yolov8-seg.hpp
include/config.h
include/utils.h
)
target_link_directories(${PROJECT_NAME} PUBLIC ${ALL_LIBS})
target_link_libraries(${PROJECT_NAME} PRIVATE nvinfer nvinfer_plugin cudart ${OpenCV_LIBS})

@ -0,0 +1,107 @@
//
// Created by ubuntu on 1/16/23.
//
#ifndef YOLOV8_TENSORRT_CSRC_SEGMENT_INCLUDE_CONFIG_H
#define YOLOV8_TENSORRT_CSRC_SEGMENT_INCLUDE_CONFIG_H
#include "opencv2/opencv.hpp"
namespace seg
{
const int DEVICE = 0;
const int INPUT_W = 640;
const int INPUT_H = 640;
const int NUM_INPUT = 1;
const int NUM_OUTPUT = 2;
const int NUM_PROPOSAL = 8400; // feature map 20*20+40*40+80*80
const int NUM_SEG_C = 32; // seg channel
const int NUM_COLS = 6 + NUM_SEG_C; // x0 y0 x1 y1 score label 32
const int SEG_W = 160;
const int SEG_H = 160;
// thresholds
const float CONF_THRES = 0.25;
const float IOU_THRES = 0.65;
const float MASK_THRES = 0.5;
// distance
const float DIS = 7680.f;
const int NUM_BINDINGS = NUM_INPUT + NUM_OUTPUT;
const cv::Scalar PAD_COLOR = { 114, 114, 114 };
const cv::Scalar RECT_COLOR = cv::Scalar(0, 0, 255);
const cv::Scalar TXT_COLOR = cv::Scalar(255, 255, 255);
const char* INPUT = "images";
const char* OUTPUT = "outputs";
const char* PROTO = "proto";
const char* CLASS_NAMES[] = {
"person", "bicycle", "car", "motorcycle", "airplane", "bus",
"train", "truck", "boat", "traffic light", "fire hydrant",
"stop sign", "parking meter", "bench", "bird", "cat",
"dog", "horse", "sheep", "cow", "elephant",
"bear", "zebra", "giraffe", "backpack", "umbrella",
"handbag", "tie", "suitcase", "frisbee", "skis",
"snowboard", "sports ball", "kite", "baseball bat", "baseball glove",
"skateboard", "surfboard", "tennis racket", "bottle", "wine glass",
"cup", "fork", "knife", "spoon", "bowl",
"banana", "apple", "sandwich", "orange", "broccoli",
"carrot", "hot dog", "pizza", "donut", "cake",
"chair", "couch", "potted plant", "bed", "dining table",
"toilet", "tv", "laptop", "mouse", "remote",
"keyboard", "cell phone", "microwave", "oven",
"toaster", "sink", "refrigerator", "book", "clock", "vase",
"scissors", "teddy bear", "hair drier", "toothbrush" };
const unsigned int COLORS[80][3] = {
{ 0, 114, 189 }, { 217, 83, 25 }, { 237, 177, 32 },
{ 126, 47, 142 }, { 119, 172, 48 }, { 77, 190, 238 },
{ 162, 20, 47 }, { 76, 76, 76 }, { 153, 153, 153 },
{ 255, 0, 0 }, { 255, 128, 0 }, { 191, 191, 0 },
{ 0, 255, 0 }, { 0, 0, 255 }, { 170, 0, 255 },
{ 85, 85, 0 }, { 85, 170, 0 }, { 85, 255, 0 },
{ 170, 85, 0 }, { 170, 170, 0 }, { 170, 255, 0 },
{ 255, 85, 0 }, { 255, 170, 0 }, { 255, 255, 0 },
{ 0, 85, 128 }, { 0, 170, 128 }, { 0, 255, 128 },
{ 85, 0, 128 }, { 85, 85, 128 }, { 85, 170, 128 },
{ 85, 255, 128 }, { 170, 0, 128 }, { 170, 85, 128 },
{ 170, 170, 128 }, { 170, 255, 128 }, { 255, 0, 128 },
{ 255, 85, 128 }, { 255, 170, 128 }, { 255, 255, 128 },
{ 0, 85, 255 }, { 0, 170, 255 }, { 0, 255, 255 },
{ 85, 0, 255 }, { 85, 85, 255 }, { 85, 170, 255 },
{ 85, 255, 255 }, { 170, 0, 255 }, { 170, 85, 255 },
{ 170, 170, 255 }, { 170, 255, 255 }, { 255, 0, 255 },
{ 255, 85, 255 }, { 255, 170, 255 }, { 85, 0, 0 },
{ 128, 0, 0 }, { 170, 0, 0 }, { 212, 0, 0 },
{ 255, 0, 0 }, { 0, 43, 0 }, { 0, 85, 0 },
{ 0, 128, 0 }, { 0, 170, 0 }, { 0, 212, 0 },
{ 0, 255, 0 }, { 0, 0, 43 }, { 0, 0, 85 },
{ 0, 0, 128 }, { 0, 0, 170 }, { 0, 0, 212 },
{ 0, 0, 255 }, { 0, 0, 0 }, { 36, 36, 36 },
{ 73, 73, 73 }, { 109, 109, 109 }, { 146, 146, 146 },
{ 182, 182, 182 }, { 219, 219, 219 }, { 0, 114, 189 },
{ 80, 183, 189 }, { 128, 128, 0 }
};
const unsigned int MASK_COLORS[20][3] = {
{ 255, 56, 56 }, { 255, 157, 151 }, { 255, 112, 31 },
{ 255, 178, 29 }, { 207, 210, 49 }, { 72, 249, 10 },
{ 146, 204, 23 }, { 61, 219, 134 }, { 26, 147, 52 },
{ 0, 212, 187 }, { 44, 153, 168 }, { 0, 194, 255 },
{ 52, 69, 147 }, { 100, 115, 255 }, { 0, 24, 236 },
{ 132, 56, 255 }, { 82, 0, 133 }, { 203, 56, 255 },
{ 255, 149, 200 }, { 255, 55, 199 }
};
struct Object
{
cv::Rect_<float> rect;
int label = 0;
float prob = 0.0;
cv::Mat boxMask;
};
}
#endif //YOLOV8_TENSORRT_CSRC_SEGMENT_INCLUDE_CONFIG_H

@ -0,0 +1,133 @@
//
// Created by ubuntu on 1/10/23.
//
#ifndef YOLOV8_CSRC_SEGMENT_INCLUDE_UTILS_H
#define YOLOV8_CSRC_SEGMENT_INCLUDE_UTILS_H
#include <sys/stat.h>
#include <iostream>
#include <string>
#include <assert.h>
#include <unistd.h>
#include "NvInfer.h"
#define CHECK(call) \
do \
{ \
const cudaError_t error_code = call; \
if (error_code != cudaSuccess) \
{ \
printf("CUDA Error:\n"); \
printf(" File: %s\n", __FILE__); \
printf(" Line: %d\n", __LINE__); \
printf(" Error code: %d\n", error_code); \
printf(" Error text: %s\n", \
cudaGetErrorString(error_code)); \
exit(1); \
} \
} while (0)
class Logger : public nvinfer1::ILogger
{
public:
nvinfer1::ILogger::Severity reportableSeverity;
explicit Logger(nvinfer1::ILogger::Severity severity = nvinfer1::ILogger::Severity::kINFO) :
reportableSeverity(severity)
{
}
void log(nvinfer1::ILogger::Severity severity, const char* msg) noexcept override
{
if (severity > reportableSeverity)
{
return;
}
switch (severity)
{
case nvinfer1::ILogger::Severity::kINTERNAL_ERROR:
std::cerr << "INTERNAL_ERROR: ";
break;
case nvinfer1::ILogger::Severity::kERROR:
std::cerr << "ERROR: ";
break;
case nvinfer1::ILogger::Severity::kWARNING:
std::cerr << "WARNING: ";
break;
case nvinfer1::ILogger::Severity::kINFO:
std::cerr << "INFO: ";
break;
default:
std::cerr << "VERBOSE: ";
break;
}
std::cerr << msg << std::endl;
}
};
inline int get_size_by_dims(const nvinfer1::Dims& dims)
{
int size = 1;
for (int i = 0; i < dims.nbDims; i++)
{
size *= dims.d[i];
}
return size;
}
inline int DataTypeToSize(const nvinfer1::DataType& dataType)
{
switch (dataType)
{
case nvinfer1::DataType::kFLOAT:
return sizeof(float);
case nvinfer1::DataType::kHALF:
return 2;
case nvinfer1::DataType::kINT8:
return sizeof(int8_t);
case nvinfer1::DataType::kINT32:
return sizeof(int32_t);
case nvinfer1::DataType::kBOOL:
return sizeof(bool);
default:
return sizeof(float);
}
}
inline float clamp(const float val, const float minVal = 0.f, const float maxVal = 1280.f)
{
assert(minVal <= maxVal);
return std::min(maxVal, std::max(minVal, val));
}
inline bool IsPathExist(const std::string& path)
{
if (access(path.c_str(), 0) == F_OK)
{
return true;
}
return false;
}
inline bool IsFile(const std::string& path)
{
if (!IsPathExist(path))
{
printf("%s:%d %s not exist\n", __FILE__, __LINE__, path.c_str());
return false;
}
struct stat buffer;
return (stat(path.c_str(), &buffer) == 0 && S_ISREG(buffer.st_mode));
}
inline bool IsFolder(const std::string& path)
{
if (!IsPathExist(path))
{
return false;
}
struct stat buffer;
return (stat(path.c_str(), &buffer) == 0 && S_ISDIR(buffer.st_mode));
}
#endif //YOLOV8_CSRC_SEGMENT_INCLUDE_UTILS_H

@ -0,0 +1,329 @@
//
// Created by ubuntu on 1/8/23.
//
#include "config.h"
#include "utils.h"
#include <fstream>
#include "NvInferPlugin.h"
using namespace seg;
class YOLOv8_seg
{
public:
explicit YOLOv8_seg(const std::string& engine_file_path);
~YOLOv8_seg();
void make_pipe(bool warmup = true);
void copy_from_Mat(const cv::Mat& image);
void infer();
void postprocess(std::vector<Object>& objs);
size_t in_size = 1 * 3 * INPUT_W * INPUT_H;
float w = INPUT_W;
float h = INPUT_H;
float ratio = 1.0f;
float dw = 0.f;
float dh = 0.f;
std::array<std::pair<int, int>, NUM_OUTPUT> out_sizes{};
std::array<void*, NUM_OUTPUT> outputs{};
private:
nvinfer1::ICudaEngine* engine = nullptr;
nvinfer1::IRuntime* runtime = nullptr;
nvinfer1::IExecutionContext* context = nullptr;
cudaStream_t stream = nullptr;
std::array<void*, NUM_BINDINGS> buffs{};
Logger gLogger{ nvinfer1::ILogger::Severity::kERROR };
};
YOLOv8_seg::YOLOv8_seg(const std::string& engine_file_path)
{
std::ifstream file(engine_file_path, std::ios::binary);
assert(file.good());
file.seekg(0, std::ios::end);
auto size = file.tellg();
std::ostringstream fmt;
file.seekg(0, std::ios::beg);
char* trtModelStream = new char[size];
assert(trtModelStream);
file.read(trtModelStream, size);
file.close();
initLibNvInferPlugins(&this->gLogger, "");
this->runtime = nvinfer1::createInferRuntime(this->gLogger);
assert(this->runtime != nullptr);
this->engine = this->runtime->deserializeCudaEngine(trtModelStream, size);
assert(this->engine != nullptr);
this->context = this->engine->createExecutionContext();
assert(this->context != nullptr);
cudaStreamCreate(&this->stream);
}
YOLOv8_seg::~YOLOv8_seg()
{
this->context->destroy();
this->engine->destroy();
this->runtime->destroy();
cudaStreamDestroy(this->stream);
for (auto& ptr : this->buffs)
{
CHECK(cudaFree(ptr));
}
for (auto& ptr : this->outputs)
{
CHECK(cudaFreeHost(ptr));
}
}
void YOLOv8_seg::make_pipe(bool warmup)
{
const nvinfer1::Dims input_dims = this->engine->getBindingDimensions(
this->engine->getBindingIndex(INPUT)
);
this->in_size = get_size_by_dims(input_dims);
CHECK(cudaMalloc(&this->buffs[0], this->in_size * sizeof(float)));
this->context->setBindingDimensions(0, input_dims);
const int32_t output_idx = this->engine->getBindingIndex(OUTPUT);
const nvinfer1::Dims output_dims = this->context->getBindingDimensions(output_idx);
this->out_sizes[output_idx - NUM_INPUT].first = get_size_by_dims(output_dims);
this->out_sizes[output_idx - NUM_INPUT].second = DataTypeToSize(
this->engine->getBindingDataType(output_idx));
const int32_t proto_idx = this->engine->getBindingIndex(PROTO);
const nvinfer1::Dims proto_dims = this->context->getBindingDimensions(proto_idx);
this->out_sizes[proto_idx - NUM_INPUT].first = get_size_by_dims(proto_dims);
this->out_sizes[proto_idx - NUM_INPUT].second = DataTypeToSize(
this->engine->getBindingDataType(proto_idx));
for (int i = 0; i < NUM_OUTPUT; i++)
{
const int osize = this->out_sizes[i].first * out_sizes[i].second;
CHECK(cudaHostAlloc(&this->outputs[i], osize, 0));
CHECK(cudaMalloc(&this->buffs[NUM_INPUT + i], osize));
}
if (warmup)
{
for (int i = 0; i < 10; i++)
{
size_t isize = this->in_size * sizeof(float);
auto* tmp = new float[isize];
CHECK(cudaMemcpyAsync(this->buffs[0],
tmp,
isize,
cudaMemcpyHostToDevice,
this->stream));
this->infer();
}
printf("model warmup 10 times\n");
}
}
void YOLOv8_seg::copy_from_Mat(const cv::Mat& image)
{
float height = (float)image.rows;
float width = (float)image.cols;
float r = std::min(INPUT_H / height, INPUT_W / width);
int padw = (int)std::round(width * r);
int padh = (int)std::round(height * r);
cv::Mat tmp;
if ((int)width != padw || (int)height != padh)
{
cv::resize(image, tmp, cv::Size(padw, padh));
}
else
{
tmp = image.clone();
}
float _dw = INPUT_W - padw;
float _dh = INPUT_H - padh;
_dw /= 2.0f;
_dh /= 2.0f;
int top = int(std::round(_dh - 0.1f));
int bottom = int(std::round(_dh + 0.1f));
int left = int(std::round(_dw - 0.1f));
int right = int(std::round(_dw + 0.1f));
cv::copyMakeBorder(tmp, tmp, top, bottom, left, right, cv::BORDER_CONSTANT, PAD_COLOR);
cv::dnn::blobFromImage(tmp,
tmp,
1 / 255.f,
cv::Size(),
cv::Scalar(0, 0, 0),
true,
false,
CV_32F);
CHECK(cudaMemcpyAsync(this->buffs[0],
tmp.ptr<float>(),
this->in_size * sizeof(float),
cudaMemcpyHostToDevice,
this->stream));
this->ratio = 1 / r;
this->dw = _dw;
this->dh = _dh;
this->w = width;
this->h = height;
}
void YOLOv8_seg::infer()
{
this->context->enqueueV2(buffs.data(), this->stream, nullptr);
for (int i = 0; i < NUM_OUTPUT; i++)
{
const int osize = this->out_sizes[i].first * out_sizes[i].second;
CHECK(cudaMemcpyAsync(this->outputs[i],
this->buffs[NUM_INPUT + i],
osize,
cudaMemcpyDeviceToHost,
this->stream));
}
cudaStreamSynchronize(this->stream);
}
void YOLOv8_seg::postprocess(std::vector<Object>& objs)
{
objs.clear();
auto* output = static_cast<float*>(this->outputs[0]); // x0 y0 x1 y1 s l *32
cv::Mat protos = cv::Mat(NUM_SEG_C, SEG_W * SEG_H, CV_32F,
static_cast<float*>(this->outputs[1]));
std::vector<int> labels;
std::vector<float> scores;
std::vector<cv::Rect> bboxes;
std::vector<cv::Mat> mask_confs;
for (int i = 0; i < NUM_PROPOSAL; i++)
{
float* ptr = output + i * NUM_COLS;
float score = *(ptr + 4);
if (score > CONF_THRES)
{
float x0 = *ptr++ - this->dw;
float y0 = *ptr++ - this->dh;
float x1 = *ptr++ - this->dw;
float y1 = *ptr++ - this->dh;
x0 = clamp(x0 * this->ratio, 0.f, this->w);
y0 = clamp(y0 * this->ratio, 0.f, this->h);
x1 = clamp(x1 * this->ratio, 0.f, this->w);
y1 = clamp(y1 * this->ratio, 0.f, this->h);
int label = *(++ptr);
cv::Mat mask_conf = cv::Mat(1, NUM_SEG_C, CV_32F, ++ptr);
mask_confs.push_back(mask_conf);
labels.push_back(label);
scores.push_back(score);
#if defined(BATCHED_NMS)
bboxes.push_back(cv::Rect_<float>(x0, y0, x1 - x0, y1 - y0));
#else
bboxes.push_back(cv::Rect_<float>(x0 + label * DIS,
y0 + label * DIS,
x1 - x0,
y1 - y0));
#endif
}
}
std::vector<int> indices;
#if defined(BATCHED_NMS)
cv::dnn::NMSBoxesBatched(bboxes, scores, labels, CONF_THRES, IOU_THRES, indices);
#else
cv::dnn::NMSBoxes(bboxes, scores, CONF_THRES, IOU_THRES, indices);
#endif
cv::Mat masks;
for (auto& i : indices)
{
#if defined(BATCHED_NMS)
cv::Rect tmp = bboxes[i];
#else
cv::Rect tmp = { (int)(bboxes[i].x - labels[i] * DIS),
(int)(bboxes[i].y - labels[i] * DIS),
bboxes[i].width,
bboxes[i].height };
#endif
Object obj;
obj.label = labels[i];
obj.rect = tmp;
obj.prob = scores[i];
masks.push_back(mask_confs[i]);
objs.push_back(obj);
}
cv::Mat matmulRes = (masks * protos).t();
cv::Mat maskMat = matmulRes.reshape(indices.size(), { SEG_W, SEG_H });
std::vector<cv::Mat> maskChannels;
cv::split(maskMat, maskChannels);
int scale_dw = this->dw / INPUT_W * SEG_W;
int scale_dh = this->dh / INPUT_H * SEG_H;
cv::Rect roi(
scale_dw,
scale_dh,
SEG_W - 2 * scale_dw,
SEG_H - 2 * scale_dh);
for (int i = 0; i < indices.size(); i++)
{
cv::Mat dest, mask;
cv::exp(-maskChannels[i], dest);
dest = 1.0 / (1.0 + dest);
dest = dest(roi);
cv::resize(dest, mask, cv::Size((int)this->w, (int)this->h), cv::INTER_LINEAR);
objs[i].boxMask = mask(objs[i].rect) > MASK_THRES;
}
}
static void draw_objects(const cv::Mat& image, cv::Mat& res, const std::vector<Object>& objs)
{
res = image.clone();
cv::Mat mask = image.clone();
for (auto& obj : objs)
{
int idx = obj.label;
cv::Scalar color = cv::Scalar(COLORS[idx][0], COLORS[idx][1], COLORS[idx][2]);
cv::Scalar mask_color = cv::Scalar(
MASK_COLORS[idx % 20][0], MASK_COLORS[idx % 20][1], MASK_COLORS[idx % 20][2]);
cv::rectangle(res, obj.rect, color, 2);
char text[256];
sprintf(text, "%s %.1f%%", CLASS_NAMES[idx], obj.prob * 100);
mask(obj.rect).setTo(mask_color, obj.boxMask);
int baseLine = 0;
cv::Size label_size = cv::getTextSize(text, cv::FONT_HERSHEY_SIMPLEX, 0.4, 1, &baseLine);
int x = (int)obj.rect.x;
int y = (int)obj.rect.y + 1;
if (y > res.rows)
y = res.rows;
cv::rectangle(res, cv::Rect(x, y, label_size.width, label_size.height + baseLine), RECT_COLOR, -1);
cv::putText(res, text, cv::Point(x, y + label_size.height),
cv::FONT_HERSHEY_SIMPLEX, 0.4, TXT_COLOR, 1);
}
cv::addWeighted(res, 0.5, mask, 0.8, 1, res);
}

@ -0,0 +1,86 @@
//
// Created by ubuntu on 1/8/23.
//
#include "include/yolov8-seg.hpp"
int main(int argc, char** argv)
{
cudaSetDevice(DEVICE);
const std::string engine_file_path{ argv[1] };
const std::string path{ argv[2] };
std::vector<cv::String> imagePathList;
bool isVideo{ false };
if (IsFile(path))
{
std::string suffix = path.substr(path.find_last_of('.') + 1);
if (suffix == "jpg")
{
imagePathList.push_back(path);
}
else if (suffix == "mp4")
{
isVideo = true;
}
}
else if (IsFolder(path))
{
cv::glob(path + "/*.jpg", imagePathList);
}
auto* yolov8 = new YOLOv8_seg(engine_file_path);
yolov8->make_pipe(true);
cv::Mat res;
cv::namedWindow("result", cv::WINDOW_AUTOSIZE);
if (isVideo)
{
cv::VideoCapture cap(path);
cv::Mat image;
if (!cap.isOpened())
{
printf("can not open ...\n");
return -1;
}
double fp_ = cap.get(cv::CAP_PROP_FPS);
int fps = round(1000.0 / fp_);
while (cap.read(image))
{
auto start = std::chrono::system_clock::now();
yolov8->copy_from_Mat(image);
yolov8->infer();
std::vector<Object> objs;
yolov8->postprocess(objs);
draw_objects(image, res, objs);
auto end = std::chrono::system_clock::now();
auto tc = std::chrono::duration_cast<std::chrono::microseconds>(end - start).count() / 1000.f;
cv::imshow("result", res);
printf("cost %2.4f ms\n", tc);
if (cv::waitKey(fps) == 'q')
{
break;
}
}
}
else
{
for (auto path : imagePathList)
{
cv::Mat image = cv::imread(path);
yolov8->copy_from_Mat(image);
auto start = std::chrono::system_clock::now();
yolov8->infer();
auto end = std::chrono::system_clock::now();
auto tc = std::chrono::duration_cast<std::chrono::microseconds>(end - start).count() / 1000.f;
printf("infer %-20s\tcost %2.4f ms\n", path.c_str(), tc);
std::vector<Object> objs;
yolov8->postprocess(objs);
draw_objects(image, res, objs);
cv::imshow("result", res);
cv::waitKey(0);
}
}
cv::destroyAllWindows();
delete yolov8;
return 0;
}

@ -53,13 +53,12 @@ def main(args):
model(fake_input) model(fake_input)
save_path = args.weights.replace('.pt', '.onnx') save_path = args.weights.replace('.pt', '.onnx')
with BytesIO() as f: with BytesIO() as f:
torch.onnx.export( torch.onnx.export(model,
model, fake_input,
fake_input, f,
f, opset_version=args.opset,
opset_version=args.opset, input_names=['images'],
input_names=['images'], output_names=['outputs', 'proto'])
output_names=['bboxes', 'scores', 'labels', 'maskconf', 'proto'])
f.seek(0) f.seek(0)
onnx_model = onnx.load(f) onnx_model = onnx.load(f)
onnx.checker.check_model(onnx_model) onnx.checker.check_model(onnx_model)

@ -102,7 +102,7 @@ def main(args):
# set desired output names order # set desired output names order
if args.seg: if args.seg:
Engine.set_desired(['bboxes', 'scores', 'labels', 'maskconf', 'proto']) Engine.set_desired(['outputs', 'proto'])
else: else:
Engine.set_desired(['num_dets', 'bboxes', 'scores', 'labels']) Engine.set_desired(['num_dets', 'bboxes', 'scores', 'labels'])
@ -180,19 +180,21 @@ def crop_mask(masks: Tensor, bboxes: Tensor) -> Tensor:
def seg_postprocess( def seg_postprocess(
data: Tuple[Tensor, Tensor, Tensor, Tensor, Tensor], data: Tuple[Tensor],
shape: Union[Tuple, List], shape: Union[Tuple, List],
conf_thres: float = 0.25, conf_thres: float = 0.25,
iou_thres: float = 0.65) -> Tuple[Tensor, Tensor, Tensor, List]: iou_thres: float = 0.65) -> Tuple[Tensor, Tensor, Tensor, List]:
assert len(data) == 5 assert len(data) == 2
h, w = shape[0] // 4, shape[1] // 4 # 4x downsampling h, w = shape[0] // 4, shape[1] // 4 # 4x downsampling
bboxes, scores, labels, maskconf, proto = (i[0] for i in data) outputs, proto = (i[0] for i in data)
bboxes, scores, labels, maskconf = outputs.split([4, 1, 1, 32], 1)
scores, labels = scores.squeeze(), labels.squeeze()
select = scores > conf_thres select = scores > conf_thres
bboxes, scores, labels, maskconf = bboxes[select], scores[select], labels[ bboxes, scores, labels, maskconf = bboxes[select], scores[select], labels[
select], maskconf[select] select], maskconf[select]
idx = batched_nms(bboxes, scores, labels, iou_thres) idx = batched_nms(bboxes, scores, labels, iou_thres)
bboxes, scores, labels, maskconf = bboxes[idx], scores[idx], labels[ bboxes, scores, labels, maskconf = bboxes[idx], scores[idx], labels[
idx], maskconf[idx] idx].int(), maskconf[idx]
masks = (maskconf @ proto).view(-1, h, w) masks = (maskconf @ proto).view(-1, h, w)
masks = crop_mask(masks, bboxes / 4.) masks = crop_mask(masks, bboxes / 4.)
masks = F.interpolate(masks[None], masks = F.interpolate(masks[None],

@ -140,7 +140,8 @@ class PostSeg(nn.Module):
[self.cv4[i](x[i]).view(bs, self.nm, -1) for i in range(self.nl)], [self.cv4[i](x[i]).view(bs, self.nm, -1) for i in range(self.nl)],
2) # mask coefficients 2) # mask coefficients
box, score, cls = self.forward_det(x) box, score, cls = self.forward_det(x)
return box, score, cls, mc.transpose(1, 2), p.flatten(2) out = torch.cat([box, score, cls, mc.transpose(1, 2)], 2)
return out, p.flatten(2)
def forward_det(self, x): def forward_det(self, x):
shape = x[0].shape shape = x[0].shape
@ -160,7 +161,7 @@ class PostSeg(nn.Module):
box0, box1 = -box[:, :2, ...], box[:, 2:, ...] box0, box1 = -box[:, :2, ...], box[:, 2:, ...]
box = self.anchors.repeat(b, 2, 1) + torch.cat([box0, box1], 1) box = self.anchors.repeat(b, 2, 1) + torch.cat([box0, box1], 1)
box = box * self.strides box = box * self.strides
score, cls = cls.transpose(1, 2).max(dim=-1) score, cls = cls.transpose(1, 2).max(dim=-1, keepdim=True)
return box.transpose(1, 2), score, cls return box.transpose(1, 2), score, cls

Loading…
Cancel
Save