darknet.dll 在C#中调用出错 cuDNN Error cudnn_convolutional_setup()

Question

我采用的时darknet V4版本，在使用CUDA10.2 CUDNN8.0.5 OPENCV4.5.3编译完毕darknet之后，做darknet.exe测试，可以检测到dog图片，采用调用摄像头检测也是可以，并且可以看到GPU已经被使用。但是在使用darknet.dll时，出现错误。加载网络到一定时就报错，具体错误代码如下：

151 route  147                                    ->   38 x  38 x 256 
 152 conv    512       3 x 3/ 2     38 x  38 x 256 ->   19 x  19 x 512 0.852 BF
 153 route  152 116                                ->   19 x  19 x1024 
 154 conv    512       1 x 1/ 1     19 x  19 x1024 ->   19 x  19 x 512 0.379 BF
 155 conv   1024       3 x 3/ 1     19 x  19 x 512 ->   19 x  19 x1024 3.407 BF
 156 conv    512       1 x 1/ 1     19 x  19 x1024 ->   19 x  19 x 512 0.379 BF
 157 conv   1024       3 x 3/ 1     19 x  19 x 512 ->   19 x  19 x1024 3.407 BF
 158 conv    512       1 x 1/ 1     19 x  19 x1024 ->   19 x  19 x 512 0.379 BF
 159 conv   1024       3 x 3/ 1     19 x  19 x 512 ->   19 x  19 x1024 3.407 BF
 160 conv    255       1 x 1/ 1     19 x  19 x1024 ->   19 x  19 x 255 0.189 BF
 161 yolo
[yolo] params: iou loss: ciou (4), iou_norm: 0.07, obj_norm: 1.00, cls_norm: 1.00, delta_norm: 1.00, scale_x_y: 1.05
Total BFLOPS 128.459 
avg_outputs = 1068395 
 Allocate additional workspace_size = 81.03 MB 
Loading weights from .\yolov4.weights... Used GPU 0 
net.optimized_memory = 0 
mini_batch = 1, batch = 16, time_steps = 1, train = 1 
Create CUDA-stream - 0 
 Create cudnn-handle 0 
nms_kind: greedynms (1), beta = 0.600000 
nms_kind: greedynms (1), beta = 0.600000 
nms_kind: greedynms (1), beta = 0.600000 
Done! Loaded 162 layers from weights-file 
Darknet error location: E:\TRY\darknet-master\src\dark_cuda.c, cudnn_check_error, line #205
cuDNN Error: CUDNN_STATUS_BAD_PARAM: Bad file descriptor
 seen 64, trained: 32032 K-images (500 Kilo-batches_64) 
 cuDNN status Error in: file: E:\TRY\darknet-master\src\convolutional_layer.c : cudnn_convolutional_setup() : line: 253 : build time: Aug 17 2021 - 20:41:18 
 cuDNN Error: CUDNN_STATUS_BAD_PARAM。

我调用darknet.dll采用的C# wrapper。其调用的代码为：

  [DllImport(YoloLibraryGpu, EntryPoint = "init")]
        internal static extern int InitializeYoloGpu(string configurationFilename, string weightsFilename, int gpuIndex);

        [DllImport(YoloLibraryGpu, EntryPoint = "detect_image")]
        internal static extern int DetectImageGpu(string filename, ref BboxContainer container);

        [DllImport(YoloLibraryGpu, EntryPoint = "detect_mat")]
        internal static extern int DetectImageGpu(IntPtr pArray, int nSize, ref BboxContainer container);

        [DllImport(YoloLibraryGpu, EntryPoint = "dispose")]
        internal static extern int DisposeYoloGpu();

        [DllImport(YoloLibraryGpu, EntryPoint = "get_device_count")]
        internal static extern int GetDeviceCount();

        [DllImport(YoloLibraryGpu, EntryPoint = "get_device_name")]
        internal static extern int GetDeviceName(int gpu, StringBuilder deviceName);
        //每次程序都在初始化darknet网络的时候报错，即在下面这一行代码：
     this._yoloWrapper = new YoloWrapper(config.ConfigFile, config.WeightsFile, config.NamesFile, gpuConfig);

后面又换到cuda11.0，cudnn换到8.2.0，问题还是照旧出现！！！！
请码友帮忙参谋一下，请收下我的膝盖！

Answer 1

暂时不知道什么原因。你可以把测试打包代码发一下，不然光看也不知道原因。我猜测是权重初始化出了问题，dark net我也常常遇到各种莫名其妙的问题。

Answer 2

YoloWrapper的代码如下：

using Alturos.Yolo.Model;
using System;
using System.Collections.Generic;
using System.IO;
using System.Linq;
using System.Runtime.InteropServices;
using System.Text;

namespace Alturos.Yolo
{
    public class YoloWrapper : IDisposable
    {
        public const int MaxObjects = 1000;
        private const string YoloLibraryCpu = "yolo_cpp_dll_cpu";
        private const string YoloLibraryGpu = "darknet";

        private readonly ImageAnalyzer _imageAnalyzer = new ImageAnalyzer();
        private readonly IYoloSystemValidator _yoloSystemValidator;
        private YoloObjectTypeResolver _objectTypeResolver;

        public DetectionSystem DetectionSystem { get; private set; } = DetectionSystem.Unknown;

        #region DllImport Cpu

        [DllImport(YoloLibraryCpu, EntryPoint = "init")]
        private static extern int InitializeYoloCpu(string configurationFilename, string weightsFilename, int gpuIndex);

        [DllImport(YoloLibraryCpu, EntryPoint = "detect_image")]
        internal static extern int DetectImageCpu(string filename, ref BboxContainer container);

        [DllImport(YoloLibraryCpu, EntryPoint = "detect_mat")]
        internal static extern int DetectImageCpu(IntPtr pArray, int nSize, ref BboxContainer container);

        [DllImport(YoloLibraryCpu, EntryPoint = "dispose")]
        internal static extern int DisposeYoloCpu();

        [DllImport(YoloLibraryCpu, EntryPoint = "built_with_opencv")]
        internal static extern bool BuiltWithOpenCV();

        #endregion

        #region DllImport Gpu

        [DllImport(YoloLibraryGpu, EntryPoint = "init")]
        internal static extern int InitializeYoloGpu(string configurationFilename, string weightsFilename, int gpuIndex);

        [DllImport(YoloLibraryGpu, EntryPoint = "detect_image")]
        internal static extern int DetectImageGpu(string filename, ref BboxContainer container);

        [DllImport(YoloLibraryGpu, EntryPoint = "detect_mat")]
        internal static extern int DetectImageGpu(IntPtr pArray, int nSize, ref BboxContainer container);

        [DllImport(YoloLibraryGpu, EntryPoint = "dispose")]
        internal static extern int DisposeYoloGpu();

        [DllImport(YoloLibraryGpu, EntryPoint = "get_device_count")]
        internal static extern int GetDeviceCount();

        [DllImport(YoloLibraryGpu, EntryPoint = "get_device_name")]
        internal static extern int GetDeviceName(int gpu, StringBuilder deviceName);

#endregion

        /// <summary>
        /// Initialize Yolo
        /// </summary>
        /// <param name="yoloConfiguration"></param>
        /// <param name="ignoreGpu">Disable automatic gpu detection</param>
        /// <exception cref="NotSupportedException">Thrown when the process not run in 64bit</exception>
        /// <exception cref="YoloInitializeException">Thrown if an error occurs during initialization</exception>
        public YoloWrapper(YoloConfiguration yoloConfiguration, GpuConfig gpuConfig = null, IYoloSystemValidator yoloSystemValidator = null)
        {
            if (yoloSystemValidator == null)
            {
                this._yoloSystemValidator = new DefaultYoloSystemValidator();
            }

            this.Initialize(yoloConfiguration.ConfigFile, yoloConfiguration.WeightsFile, yoloConfiguration.NamesFile, gpuConfig);
        }

        /// <summary>
        /// Initialize Yolo
        /// </summary>
        /// <param name="configurationFilename">Yolo configuration (.cfg) file path</param>
        /// <param name="weightsFilename">Yolo trainded data (.weights) file path</param>
        /// <param name="namesFilename">Yolo object names (.names) file path</param>
        /// <param name="gpu">Gpu Index if multiple graphic devices available</param>
        /// <param name="ignoreGpu">Disable automatic gpu detection</param>
        /// <exception cref="NotSupportedException">Thrown when the process not run in 64bit</exception>
        /// <exception cref="YoloInitializeException">Thrown if an error occurs during initialization</exception>
        public YoloWrapper(string configurationFilename, string weightsFilename, string namesFilename, GpuConfig gpuConfig = null, IYoloSystemValidator yoloSystemValidator = null)
        {
            if (yoloSystemValidator == null)
            {
                this._yoloSystemValidator = new DefaultYoloSystemValidator();
            }

            this.Initialize(configurationFilename, weightsFilename, namesFilename, gpuConfig);
        }

        public void Dispose()
        {
            switch (this.DetectionSystem)
            {
                case DetectionSystem.CPU:
                    DisposeYoloCpu();
                    break;
                case DetectionSystem.GPU:
                    DisposeYoloGpu();
                    break;
            }
        }

        private void Initialize(string configurationFilename, string weightsFilename, string namesFilename, GpuConfig gpuConfig)
        {
            if (IntPtr.Size != 8)
            {
                throw new NotSupportedException("Only 64-bit processes are supported");
            }

            var systemReport = this._yoloSystemValidator.Validate();
            if (!systemReport.MicrosoftVisualCPlusPlusRedistributableExists)
            {
                throw new YoloInitializeException("Microsoft Visual C++ 2017-2019 Redistributable (x64)");
            }

            this.DetectionSystem = DetectionSystem.CPU;

            if (gpuConfig != null)
            {
                if (!systemReport.CudaExists)
                {
                    throw new YoloInitializeException("Cuda files not found");
                }

                if (!systemReport.CudnnExists)
                {
                    throw new YoloInitializeException("Cudnn not found");
                }

                var deviceCount = GetDeviceCount();
                if (deviceCount == 0)
                {
                    throw new YoloInitializeException("No Nvidia graphic device is available");
                }

                if (gpuConfig.GpuIndex > (deviceCount - 1))
                {
                    throw new YoloInitializeException("Graphic device index is out of range");
                }

                this.DetectionSystem = DetectionSystem.GPU;
            }

            switch (this.DetectionSystem)
            {
                case DetectionSystem.CPU:
                    InitializeYoloCpu(configurationFilename, weightsFilename, 0);
                    break;
                case DetectionSystem.GPU:
                    InitializeYoloGpu(configurationFilename, weightsFilename, gpuConfig.GpuIndex);
                    break;
            }

            this._objectTypeResolver = new YoloObjectTypeResolver(namesFilename);
        }

        /// <summary>
        /// Detect objects on an image
        /// </summary>
        /// <param name="filepath"></param>
        /// <returns></returns>
        /// <exception cref="FileNotFoundException">Thrown when the filepath is wrong</exception>
        public IEnumerable<YoloItem> Detect(string filepath)
        {
            if (!File.Exists(filepath))
            {
                throw new FileNotFoundException("Cannot find the file", filepath);
            }

            var container = new BboxContainer();
            var count = 0;
            switch (this.DetectionSystem)
            {
                case DetectionSystem.CPU:
                    count = DetectImageCpu(filepath, ref container);
                    break;
                case DetectionSystem.GPU:
                    count = DetectImageGpu(filepath, ref container);
                    break;
            }

            if (count == -1)
            {
                throw new NotImplementedException("C++ dll compiled incorrectly");
            }

            return this.Convert(container);
        }

        /// <summary>
        /// Detect objects on an image
        /// </summary>
        /// <param name="imageData"></param>
        /// <returns></returns>
        /// <exception cref="NotImplementedException">Thrown when the yolo_cpp dll is wrong compiled</exception>
        /// <exception cref="Exception">Thrown when the byte array is not a valid image</exception>
        public IEnumerable<YoloItem> Detect(byte[] imageData)
        {
            if (!this._imageAnalyzer.IsValidImageFormat(imageData))
            {
                throw new Exception("Invalid image data, wrong image format");
            }

            var container = new BboxContainer();
            var size = Marshal.SizeOf(imageData[0]) * imageData.Length;
            var pnt = Marshal.AllocHGlobal(size);

            var count = 0;
            try
            {
                // Copy the array to unmanaged memory.
                Marshal.Copy(imageData, 0, pnt, imageData.Length);
                switch (this.DetectionSystem)
                {
                    case DetectionSystem.CPU:
                        count = DetectImageCpu(pnt, imageData.Length, ref container);
                        break;
                    case DetectionSystem.GPU:
                        count = DetectImageGpu(pnt, imageData.Length, ref container);
                        break;
                }
            }
            catch (Exception)
            {
                return null;
            }
            finally
            {
                // Free the unmanaged memory.
                Marshal.FreeHGlobal(pnt);
            }

            if (count == -1)
            {
                throw new NotImplementedException("C++ dll compiled incorrectly");
            }

            return this.Convert(container);
        }

        /// <summary>
        /// Detect objects on an image
        /// </summary>
        /// <param name="imagePtr"></param>
        /// <param name="size"></param>
        /// <returns></returns>
        /// <exception cref="NotImplementedException">Thrown when the yolo_cpp dll is wrong compiled</exception>
        public IEnumerable<YoloItem> Detect(IntPtr imagePtr, int size)
        {
            var container = new BboxContainer();

            var count = 0;
            try
            {
                switch (this.DetectionSystem)
                {
                    case DetectionSystem.CPU:
                        count = DetectImageCpu(imagePtr, size, ref container);
                        break;
                    case DetectionSystem.GPU:
                        count = DetectImageGpu(imagePtr, size, ref container);
                        break;
                }
            }
            catch (Exception)
            {
                return null;
            }

            if (count == -1)
            {
                throw new NotImplementedException("C++ dll compiled incorrectly");
            }

            return this.Convert(container);
        }

        public string GetGraphicDeviceName(GpuConfig gpuConfig)
        {
            if (gpuConfig == null)
            {
                return string.Empty;
            }

            var systemReport = this._yoloSystemValidator.Validate();
            if (!systemReport.CudaExists || !systemReport.CudnnExists)
            {
                return "unknown";
            }

            var deviceName = new StringBuilder(); //allocate memory for string
            GetDeviceName(gpuConfig.GpuIndex, deviceName);
            return deviceName.ToString();
        }

        public bool IsBuiltWithOpenCV()
        {
            return BuiltWithOpenCV();
        }

        private IEnumerable<YoloItem> Convert(BboxContainer container)
        {
            return container.candidates.Where(o => o.h > 0 || o.w > 0).Select(o =>

                new YoloItem
                {
                    X = (int)o.x,
                    Y = (int)o.y,
                    Height = (int)o.h,
                    Width = (int)o.w,
                    Confidence = o.prob,
                    Type = this._objectTypeResolver.Resolve((int)o.obj_id)
                }
            );
        }
    }
}

调用的代码如下：

 private void Initialize(YoloConfiguration config)
        {
            try
            {
                if (this._yoloWrapper != null)
                {
                    this._yoloWrapper.Dispose();
                }

                var gpuConfig = new GpuConfig();
                // var useOnlyCpu = this.cpuToolStripMenuItem.Checked;
                bool useOnlyCpu = false;
                if (useOnlyCpu)
                {
                    gpuConfig = null;
                }

                this.toolStripStatusLabelYoloInfo.Text = $"Initialize...";

                var sw = new Stopwatch();
                sw.Start();
                this._yoloWrapper = new YoloWrapper(config.ConfigFile, config.WeightsFile, config.NamesFile, gpuConfig);
                sw.Stop();

                var action = new MethodInvoker(delegate ()
                {
                    var deviceName = this._yoloWrapper.GetGraphicDeviceName(gpuConfig);
                    this.toolStripStatusLabelYoloInfo.Text = $"Initialize Yolo in {sw.Elapsed.TotalMilliseconds:0} ms - Detection System:{this._yoloWrapper.DetectionSystem} {deviceName} Weights:{config.WeightsFile}";
                });

                this.statusStrip1.Invoke(action);
                this.buttonProcessImage.Invoke(new MethodInvoker(delegate () { this.buttonProcessImage.Enabled = true; }));
                this.buttonStartTracking.Invoke(new MethodInvoker(delegate () { this.buttonStartTracking.Enabled = true; }));
            }
            catch (Exception exception)
            {
                MessageBox.Show($"{nameof(Initialize)} - {exception}", "Error Initialize", MessageBoxButtons.OK, MessageBoxIcon.Error);
            }
        }        

        private void DetectSelectedImage()
        {
            var items = this.Detect();
            this.dataGridViewResult.DataSource = items;
            this.DrawBoundingBoxes(items);
        }

        private List<YoloItem> Detect(bool memoryTransfer = true)
        {
            if (this._yoloWrapper == null)
            {
                return null;
            }

            var imageInfo = this.GetCurrentImage();
            var imageData = File.ReadAllBytes(imageInfo.Path);

            var sw = new Stopwatch();
            sw.Start();
            List<YoloItem> items;
            if (memoryTransfer)
            {
                items = this._yoloWrapper.Detect(imageData).ToList();
            }
            else
            {
                items = this._yoloWrapper.Detect(imageInfo.Path).ToList();
            }
            sw.Stop();
            this.groupBoxResult.Text = $"Result [ processed in {sw.Elapsed.TotalMilliseconds:0} ms ]";

            return items;
        }

每出错是在 this._yoloWrapper = new YoloWrapper(config.ConfigFile, config.WeightsFile, config.NamesFile, gpuConfig);这段代码处，一旦带GPU初始化，就出错，用CPU初始化就可能通过

Answer 3

cfg结构如下：
[net]

Testing

#batch=1
#subdivisions=1

Training

batch=64
subdivisions=16
width=416
height=416
channels=3
momentum=0.9
decay=0.0005
angle=0
saturation = 1.5
exposure = 1.5
hue=.1

learning_rate=0.001
burn_in=1000
max_batches = 500200
policy=steps
steps=400000,450000
scales=.1,.1

[convolutional]
batch_normalize=1
filters=32
size=3
stride=1
pad=1
activation=leaky

Downsample

[convolutional]
batch_normalize=1
filters=64
size=3
stride=2
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=32
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=64
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

Downsample

[convolutional]
batch_normalize=1
filters=128
size=3
stride=2
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=64
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=128
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

[convolutional]
batch_normalize=1
filters=64
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=128
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

Downsample

[convolutional]
batch_normalize=1
filters=256
size=3
stride=2
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

Downsample

[convolutional]
batch_normalize=1
filters=512
size=3
stride=2
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

Downsample

[convolutional]
batch_normalize=1
filters=1024
size=3
stride=2
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=512
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=1024
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

[convolutional]
batch_normalize=1
filters=512
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=1024
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

[convolutional]
batch_normalize=1
filters=512
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=1024
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

[convolutional]
batch_normalize=1
filters=512
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
filters=1024
size=3
stride=1
pad=1
activation=leaky

[shortcut]
from=-3
activation=linear

######################

[convolutional]
batch_normalize=1
filters=512
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
size=3
stride=1
pad=1
filters=1024
activation=leaky

[convolutional]
batch_normalize=1
filters=512
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
size=3
stride=1
pad=1
filters=1024
activation=leaky

[convolutional]
batch_normalize=1
filters=512
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
size=3
stride=1
pad=1
filters=1024
activation=leaky

[convolutional]
size=1
stride=1
pad=1
filters=255
activation=linear

[yolo]
mask = 6,7,8
anchors = 10,13, 16,30, 33,23, 30,61, 62,45, 59,119, 116,90, 156,198, 373,326
classes=80
num=9
jitter=.3
ignore_thresh = .7
truth_thresh = 1
random=1

[route]
layers = -4

[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky

[upsample]
stride=2

[route]
layers = -1, 61

[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
size=3
stride=1
pad=1
filters=512
activation=leaky

[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
size=3
stride=1
pad=1
filters=512
activation=leaky

[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
size=3
stride=1
pad=1
filters=512
activation=leaky

[convolutional]
size=1
stride=1
pad=1
filters=255
activation=linear

[yolo]
mask = 3,4,5
anchors = 10,13, 16,30, 33,23, 30,61, 62,45, 59,119, 116,90, 156,198, 373,326
classes=80
num=9
jitter=.3
ignore_thresh = .7
truth_thresh = 1
random=1

[route]
layers = -4

[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky

[upsample]
stride=2

[route]
layers = -1, 36

[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
size=3
stride=1
pad=1
filters=256
activation=leaky

[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
size=3
stride=1
pad=1
filters=256
activation=leaky

[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky

[convolutional]
batch_normalize=1
size=3
stride=1
pad=1
filters=256
activation=leaky

[convolutional]
size=1
stride=1
pad=1
filters=255
activation=linear

[yolo]
mask = 0,1,2
anchors = 10,13, 16,30, 33,23, 30,61, 62,45, 59,119, 116,90, 156,198, 373,326
classes=80
num=9
jitter=.3
ignore_thresh = .7
truth_thresh = 1
random=1

Answer 4

YOLOv3 darknet启用cudnn编译失败_蔓^o^的博客-CSDN博客 yolo@TOC欢迎使用Markdown编辑器你好！ 这是你第一次使用 Markdown编辑器 所展示的欢迎页。如果你想学习如何使用Markdown编辑器, 可以仔细阅读这篇文章，了解一下Markdown的基本语法知识。新的改变我们对Markdown编辑器进行了一些功能拓展与语法支持，除了标准的Markdown编辑器功能，我们增加了如下几点新功能，帮助你用它写博客：全新的界面设计 ，将... https://blog.csdn.net/weixin_43738167/article/details/93488840?ops_request_misc=&request_id=&biz_id=102&utm_term=darknet.dll%20%E5%9C%A8C?ops_request_misc=&request_id=&biz_id=102&utm_term=darknet.dll%20%E5%9C%A8C&utm_medium=distribute.pc_search_result.none-task-blog-2~all~sobaiduweb~default-0-.first_rank_v2_pc_rank_v29&spm=1018.2226.3001.4187#%E4%B8%AD%E8%B0%83%E7%94%A8%E5%87%BA%E9%94%99%20cuDNN%20Err&utm_medium=distribute.pc_search_result.none-task-blog-2~all~sobaiduweb~default-0-.first_rank_v2_pc_rank_v29

Answer 5

尝试了以上版本的OPENCV

Answer 6

c++ - OpenCV 4.x+ requires enabled C++11 support compilation darknet fatal error - Stack Overflow I just compiled and installed the last version of OpenCV 3.4.0 and I would like to compile darknet (for yolo object detection), but at compilation, I have gcc -Iinclude/ -Isrc/ -DOPENCV `pkg-confi... https://stackoverflow.com/questions/49943161/opencv-4-x-requires-enabled-c11-support-compilation-darknet-fatal-error
compile opencv with darknet wrong #691
compile opencv with darknet wrong · Issue #691 · pjreddie/darknet · GitHub hi,when I compile opencv with darknet,the following error occurred: gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -Wall -Wno-unused-result -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast ... https://github.com/pjreddie/darknet/issues/691

Answer 7

我的编译环境是WIN10+VS2019 显卡为GF2070

Answer 8

GF2070

Answer 9

望采纳，谢谢。

1D Dilated convolution Descriptor setup - cuDNN - NVIDIA Developer Forums If anyone could share some wisdom with me that would be great. I’m coding a 1D timeseries NN with dilated convolutional layers. I can’t seem to find a working set of descriptors for these dilated convolutional layers. I’… https://forums.developer.nvidia.com/t/1d-dilated-convolution-descriptor-setup/178545

https://github.com/facebookarchive/fb.resnet.torch/issues/139

Answer 10

问题还是出现啊！

Answer 11

上面有两个链接，应该可以解决

Answer 12

darknet编译出现的问题_darren2015zdc的博客-CSDN博客 darknet的git: https://github.com/pjreddie/darknet/tree/master一般训练和图片的推测， 只需要编译 cuda, cudnn 就可以了。但是在使用 opencv的时候，对应的 cuda版本会有问题。opencv3.4.0 和cuda9 以下 可以正常编译。问题1. cuda版本与opencv版本冲突CMake Error: The f... https://blog.csdn.net/darren2015zdc/article/details/91492043?ops_request_misc=&request_id=&biz_id=102&utm_term=darknet.dll%20%E5%9C%A8C?ops_request_misc=&request_id=&biz_id=102&utm_term=darknet.dll%20%E5%9C%A8C&utm_medium=distribute.pc_search_result.none-task-blog-2~all~sobaiduweb~default-5-.first_rank_v2_pc_rank_v29&spm=1018.2226.3001.4187#%E4%B8%AD%E8%B0%83%E7%94%A8%E5%87%BA%E9%94%99%20cuDNN%20Err&utm_medium=distribute.pc_search_result.none-task-blog-2~all~sobaiduweb~default-5-.first_rank_v2_pc_rank_v29

Answer 13

@laust9 你是什么语言

Answer 14

c#

Answer 15

我也遇到相同问题 请问你解决了吗