jetson-inference
jetson-inference copied to clipboard
dusty-nv
Didn't load class colours...
My model trained for two classes object-1 and object-goal
I trained the ssd-mobilenet-v2 in pytorch-ssd repository.
When I train a model, I get label.txt file with content BACKGROUND, object-1 and, object-goal
But when I try to load the model for predicting, it gives me the error:
I don't know why... 🤔 The classes of the model is ok...
my labels.txt file is:
BACKGROUND
object-1
object-goal
and the console output:
jetson.utils -- compiled without NumPy array conversion support (warning)
jetson.utils -- if you wish to have support for converting NumPy arrays,
jetson.utils -- first run 'sudo apt-get install python-numpy python3-numpy'
--Finding Camera Instance
--Set Grab Continousely
--Convert Pixel Format
jetson.inference -- detectNet loading custom model '1080F.onnx'
detectNet -- loading detection network model from:
-- prototxt NULL
-- model 1080F.onnx
-- input_blob 'input_0'
-- output_cvg 'scores'
-- output_bbox 'boxes'
-- mean_pixel 0.000000
-- class_labels ./labels.txt
-- class_colors NULL
-- threshold 0.500000
-- batch_size 1
[TRT] TensorRT version 8.2.1
[TRT] loading NVIDIA plugins...
[TRT] Registered plugin creator - ::GridAnchor_TRT version 1
[TRT] Registered plugin creator - ::GridAnchorRect_TRT version 1
[TRT] Registered plugin creator - ::NMS_TRT version 1
[TRT] Registered plugin creator - ::Reorg_TRT version 1
[TRT] Registered plugin creator - ::Region_TRT version 1
[TRT] Registered plugin creator - ::Clip_TRT version 1
[TRT] Registered plugin creator - ::LReLU_TRT version 1
[TRT] Registered plugin creator - ::PriorBox_TRT version 1
[TRT] Registered plugin creator - ::Normalize_TRT version 1
[TRT] Registered plugin creator - ::ScatterND version 1
[TRT] Registered plugin creator - ::RPROI_TRT version 1
[TRT] Registered plugin creator - ::BatchedNMS_TRT version 1
[TRT] Registered plugin creator - ::BatchedNMSDynamic_TRT version 1
[TRT] Could not register plugin creator - ::FlattenConcat_TRT version 1
[TRT] Registered plugin creator - ::CropAndResize version 1
[TRT] Registered plugin creator - ::DetectionLayer_TRT version 1
[TRT] Registered plugin creator - ::EfficientNMS_TRT version 1
[TRT] Registered plugin creator - ::EfficientNMS_ONNX_TRT version 1
[TRT] Registered plugin creator - ::EfficientNMS_TFTRT_TRT version 1
[TRT] Registered plugin creator - ::Proposal version 1
[TRT] Registered plugin creator - ::ProposalLayer_TRT version 1
[TRT] Registered plugin creator - ::PyramidROIAlign_TRT version 1
[TRT] Registered plugin creator - ::ResizeNearest_TRT version 1
[TRT] Registered plugin creator - ::Split version 1
[TRT] Registered plugin creator - ::SpecialSlice_TRT version 1
[TRT] Registered plugin creator - ::InstanceNormalization_TRT version 1
[TRT] detected model format - ONNX (extension '.onnx')
[TRT] desired precision specified for GPU: FASTEST
[TRT] requested fasted precision for device GPU without providing valid calibrator, disabling INT8
[TRT] [MemUsageChange] Init CUDA: CPU +225, GPU +0, now: CPU 330, GPU 3242 (MiB)
[TRT] [MemUsageSnapshot] Begin constructing builder kernel library: CPU 330 MiB, GPU 3271 MiB
[TRT] [MemUsageSnapshot] End constructing builder kernel library: CPU 359 MiB, GPU 3303 MiB
[TRT] native precisions detected for GPU: FP32, FP16
[TRT] selecting fastest native precision for GPU: FP16
[TRT] found engine cache file 1080F.onnx.1.1.8201.GPU.FP16.engine
[TRT] found model checksum 1080F.onnx.sha256sum
[TRT] echo "$(cat 1080F.onnx.sha256sum) 1080F.onnx" | sha256sum --check --status
[TRT] model matched checksum 1080F.onnx.sha256sum
[TRT] loading network plan from engine cache... 1080F.onnx.1.1.8201.GPU.FP16.engine
[TRT] device GPU, loaded 1080F.onnx
[TRT] [MemUsageChange] Init CUDA: CPU +0, GPU +0, now: CPU 347, GPU 3362 (MiB)
[TRT] Loaded engine size: 16 MiB
[TRT] Using cublas as a tactic source
[TRT] [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +157, GPU +37, now: CPU 505, GPU 3417 (MiB)
[TRT] Using cuDNN as a tactic source
[TRT] [MemUsageChange] Init cuDNN: CPU +241, GPU +17, now: CPU 746, GPU 3434 (MiB)
[TRT] Deserialization required 4799981 microseconds.
[TRT] [MemUsageChange] TensorRT-managed allocation in engine deserialization: CPU +0, GPU +16, now: CPU 0, GPU 16 (MiB)
[TRT] Using cublas as a tactic source
[TRT] [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +0, GPU +0, now: CPU 746, GPU 3434 (MiB)
[TRT] Using cuDNN as a tactic source
[TRT] [MemUsageChange] Init cuDNN: CPU +0, GPU +0, now: CPU 746, GPU 3434 (MiB)
[TRT] Total per-runner device persistent memory is 15163904
[TRT] Total per-runner host persistent memory is 70080
[TRT] Allocated activation device memory of size 8871424
[TRT] [MemUsageChange] TensorRT-managed allocation in IExecutionContext creation: CPU +0, GPU +23, now: CPU 0, GPU 39 (MiB)
[TRT]
[TRT] CUDA engine context initialized on device GPU:
[TRT] -- layers 109
[TRT] -- maxBatchSize 1
[TRT] -- deviceMemory 8871424
[TRT] -- bindings 3
[TRT] binding 0
-- index 0
-- name 'input_0'
-- type FP32
-- in/out INPUT
-- # dims 4
-- dim #0 1
-- dim #1 3
-- dim #2 300
-- dim #3 300
[TRT] binding 1
-- index 1
-- name 'scores'
-- type FP32
-- in/out OUTPUT
-- # dims 3
-- dim #0 1
-- dim #1 3000
-- dim #2 3
[TRT] binding 2
-- index 2
-- name 'boxes'
-- type FP32
-- in/out OUTPUT
-- # dims 3
-- dim #0 1
-- dim #1 3000
-- dim #2 4
[TRT]
[TRT] binding to input 0 input_0 binding index: 0
[TRT] binding to input 0 input_0 dims (b=1 c=3 h=300 w=300) size=1080000
[TRT] binding to output 0 scores binding index: 1
[TRT] binding to output 0 scores dims (b=1 c=3000 h=3 w=1) size=36000
[TRT] binding to output 1 boxes binding index: 2
[TRT] binding to output 1 boxes dims (b=1 c=3000 h=4 w=1) size=48000
[TRT]
[TRT] device GPU, 1080F.onnx initialized.
[TRT] detectNet -- number of object classes: 3
[TRT] detectNet -- maximum bounding boxes: 3000
[TRT] loaded 3 class labels
[TRT] detectNet -- number of object classes: 3
[TRT] loaded 0 class colors
[TRT] didn't load expected number of class colors (0 of 3)
[TRT] filling in remaining 3 class colors with default colors
Segmentation fault (core dumped)
I don't know what is the problem... and this my code:
'''
A simple Program for grabing video from basler camera and converting it to opencv img.
Tested on Basler acA1300-200uc (USB3, linux 64bit , python 3.5)
'''
from pypylon import pylon
import cv2,time
import torch
import yaml
import sys
import argparse
from jetson_inference import detectNet
from jetson_utils import videoSource, videoOutput, Log
from jetson_utils import cudaImage, cudaMemcpy
# parse the command line
parser = argparse.ArgumentParser(description="Locate objects in a live camera stream using an object detection DNN.",
formatter_class=argparse.RawTextHelpFormatter,
epilog=detectNet.Usage() + videoSource.Usage() + videoOutput.Usage() + Log.Usage())
parser.add_argument("input", type=str, default="", nargs='?', help="URI of the input stream")
parser.add_argument("output", type=str, default="", nargs='?', help="URI of the output stream")
parser.add_argument("--network", type=str, default="ssd-mobilenet-v2", help="pre-trained model to load (see below for options)")
parser.add_argument("--overlay", type=str, default="box,labels,conf", help="detection overlay flags (e.g. --overlay=box,labels,conf)\nvalid combinations are: 'box', 'labels', 'conf', 'none'")
parser.add_argument("--threshold", type=float, default=0.5, help="minimum detection threshold to use")
try:
args = parser.parse_known_args()[0]
except:
print("")
parser.print_help()
sys.exit(0)
# conecting to the first available camera
print("--Finding Camera Instance")
camera = pylon.InstantCamera(pylon.TlFactory.GetInstance().CreateFirstDevice())
# Grabing Continusely (video) with minimal delay
print("--Set Grab Continousely")
camera.StartGrabbing(pylon.GrabStrategy_LatestImageOnly)
converter = pylon.ImageFormatConverter()
# converting to opencv bgr format
print("--Convert Pixel Format")
converter.OutputPixelFormat = pylon.PixelType_BGR8packed
converter.OutputBitAlignment = pylon.OutputBitAlignment_MsbAligned
model = detectNet(model="1080F.onnx",labels="./labels.txt",
input_blob="input_0",output_cvg="scores" , output_bbox="boxes",
threshold=args.threshold)
while camera.IsGrabbing():
grabResult = camera.RetrieveResult(5000, pylon.TimeoutHandling_ThrowException)
if grabResult.GrabSucceeded():
# Access the image data
image1 = converter.Convert(grabResult)
image = image1.GetArray()
image_rsz = cv2.resize(image,(640,640))
image = cudaImage(width=640,height=640,like=image_rsz,format="rgb8")
image_final = cudaMemcpy(image , image_rsz)
results = model.Detect(image_final , overlay=args.overlay)
for res in results:
print(res)
cv2.namedWindow('title', cv2.WINDOW_NORMAL)
cv2.imshow('title', image_final)
k = cv2.waitKey(1)
if k == 27:
break
grabResult.Release()
# Releasing the resource
camera.StopGrabbing()
cv2.destroyAllWindows()
@dusty-nv can you help me dusty ? :(
@dusty-nv I noticed something... the model load successfully and when trying to create cuda image... it give me a error segmentation fault (core dumped) this is the problem... I don't know why Am I convert the image to cudaImage correctly ?