Currently I'm detecting lanes using tools/detect.py.

For Condlane inference, I changed this

batch_size=1 # from 8 (for condlane inference)

And tried these configs for FHD input image

img_height = 1080 # from 320
img_width = 1920 # from 800

ori_img_h = 1080 # from 590
ori_img_w = 1920 # from 1640

crop_bbox = [0,540,1920,1080] # from [0, 270, 1640, 590]

Changing img_scale = (800,320) results

The size of tensor a must match the size of tensor b at non-singleton dimension 3

How can I properly change the input image size (ex. FHD) on CondLane config file?

Hello, I would like to know if there is any way to get real-time lane detection and curvature detection using deep learning. I have seen traditional computer vision algorithms but I am looking for a Deep Learning model that could help me out with this. Any suggestions will be very helpful. Thanks in advance.

The inference outputs from the model are really bad even for very easy images.

1. Using Laneatt_Res18_Culane

2. Using SCNN_Res50_Culane

Any idea why this is happening? I've just done normal inference without any changes.

When I run: python tools/detect.py configs/resa/resa34_culane.py --img images --load_from resa_r34_culane.pth --savedir ./vis Traceback (most recent call last): File "D:/XXX/XXX/XXX/lanedet-main/tools/detect.py", line 8, in from lanedet.datasets.process import Process File "D:\XXX\XXX\XXX\lanedet-main\lanedet_init_.py", line 1, in from .ops import * File "D:\XXX\XXX\XXX\lanedet-main\lanedet\ops_init_.py", line 1, in from .nms import nms File "D:\XXX\XXX\XXX\lanedet-main\lanedet\ops\nms.py", line 29, in from . import nms_impl ImportError: cannot import name 'nms_impl' from partially initialized module 'lanedet.ops' (most likely due to a circular import) (D:\XXX\XXX\XXX\lanedet-main\lanedet\ops_init_.py)

Hello,

I have tried testing with the CULane dataset with rsea and it is working well with the example video_example/05081544_0305/
With the following image configuration: img_height = 288 img_width = 800 cut_height = 240 ori_img_h = 590 ori_img_w = 1640

But with custom image of configurations: img_height = 288 img_width = 800 cut_height = 240 ori_img_h = 1208 // 590 ori_img_w = 1920 //1640

With above parameters: custom image

With defaut parameters: custom image

Could you please assist me which params needs to be tuned.

Appreciate any response.

Regards, Ajay

`sample_input = torch.rand((32, 3, 3, 3))

torch.onnx.export( net1.module, # PyTorch Model sample_input, # Input tensor '/content/drive/MyDrive/MobileNetV2-model-onnx.onnx', # Output file (eg. 'output_model.onnx') opset_version = 12, # Operator support version input_names = ['input'], # Input tensor name (arbitary) output_names = ['output'] # Output tensor name (arbitary) )`

Got this Error:

TypeError Traceback (most recent call last) in () 5 opset_version=12, # Operator support version 6 input_names=['input'], # Input tensor name (arbitary) ----> 7 output_names=['output'] # Output tensor name (arbitary) 8 )

 21     def forward(self, batch):
 22         output = {}

---> 23 fea = self.backbone(batch['img']) 24 25 if self.aggregator:

TypeError: new(): invalid data type 'str'

from . import nms_impl ImportError: cannot import name 'nms_impl' from partially initialized module 'la nedet.ops' (most likely due to a circular import) (D:\lanedet-main\lanedet\ops_ init_.py)

Hello, Thank you so much for sharing a very much useful repository.

I have followed the step by step instructions given, and have downloaded all the datasets as mentioned in the below image

Training: python main.py configs/resa/resa50_culane.py --gpus 0

After running the above command, i was able to see following window:

But i couldn't find any model file such as culane_resnet50.pth ,resa_r34_culane.pth !! As it mentioned in the example run case.

Alternatively, is it possible to share the pre-trained model file?

As I am a beginner, I greatly appreciate your understanding and kind response.

Regards, Ajay

python setup.py build develop

File "/home/zzj/anaconda3/envs/Lanedet/lib/python3.8/site-packages/pkg_resources/_vendor/packaging/version.py", line 275, in init match = self._regex.search(version) TypeError: expected string or bytes-like object

ubuntu20.04 what can i do？

if don't modify (from .nms import nms) from lanedet/ops/init.py to (from . import *) there will be an error. and if don't modify (from . import nms_impl) from lanedet/ops/nms.py to (from . import *) there will be an error. And when run inference, there is no lanedet directory in the tools directory, resulting in module error from lanedet/tools/detect.py line 8~12. Is there any other way to remove the error?

Hey @Turoad, thanks for your work, it's very useful. I recently customized to train condlane with mobilenetv2 backbone but got this error!!

Traceback (most recent call last):
  File "main.py", line 65, in <module>
    main()
  File "main.py", line 35, in main
    runner.train()
  File "/mnt/09a762a6-3f6e-469b-8d6d-e9fa625e24b9/USER/LuanDD/lanedet/lanedet/engine/runner.py", line 94, in train
    self.train_epoch(epoch, train_loader)
  File "/mnt/09a762a6-3f6e-469b-8d6d-e9fa625e24b9/USER/LuanDD/lanedet/lanedet/engine/runner.py", line 67, in train_epoch
    output = self.net(data)
  File "/mnt/09a762a6-3f6e-469b-8d6d-e9fa625e24b9/USER/LuanDD/pyenv/lib/python3.6/site-packages/torch/nn/modules/module.py", line 889, in _call_impl
    result = self.forward(*input, **kwargs)
  File "/mnt/09a762a6-3f6e-469b-8d6d-e9fa625e24b9/USER/LuanDD/pyenv/lib/python3.6/site-packages/mmcv/parallel/data_parallel.py", line 42, in forward
    return super().forward(*inputs, **kwargs)
  File "/mnt/09a762a6-3f6e-469b-8d6d-e9fa625e24b9/USER/LuanDD/pyenv/lib/python3.6/site-packages/torch/nn/parallel/data_parallel.py", line 165, in forward
    return self.module(*inputs[0], **kwargs[0])
  File "/mnt/09a762a6-3f6e-469b-8d6d-e9fa625e24b9/USER/LuanDD/pyenv/lib/python3.6/site-packages/torch/nn/modules/module.py", line 889, in _call_impl
    result = self.forward(*input, **kwargs)
  File "/mnt/09a762a6-3f6e-469b-8d6d-e9fa625e24b9/USER/LuanDD/lanedet/lanedet/models/nets/detector.py", line 29, in forward
    fea = self.neck(fea)
  File "/mnt/09a762a6-3f6e-469b-8d6d-e9fa625e24b9/USER/LuanDD/pyenv/lib/python3.6/site-packages/torch/nn/modules/module.py", line 889, in _call_impl
    result = self.forward(*input, **kwargs)
  File "/mnt/09a762a6-3f6e-469b-8d6d-e9fa625e24b9/USER/LuanDD/lanedet/lanedet/models/necks/fpn.py", line 113, in forward
    assert len(inputs) >= len(self.in_channels)
AssertionError

Can you help me clarify it? This is my config

net = dict(
    type='Detector',
)

backbone = dict(
    type='MobileNet',
    net='MobileNetV2',
    pretrained=True,
    # replace_stride_with_dilation=[False, False, False],
    out_conv=False,
    # in_channels=[64, 128, 256, 512]
)

featuremap_out_channel = 1280
featuremap_out_stride = 32 

sample_y = range(590, 270, -8)

batch_size = 8
aggregator = dict(
    type='TransConvEncoderModule',
    in_dim=1280,
    attn_in_dims=[1280, 64],
    attn_out_dims=[64, 64],
    strides=[1, 1],
    ratios=[4, 4],
    pos_shape=(batch_size, 10, 25),
)

neck=dict(
    type='FPN',
    in_channels=[64, 128, 256, 64],
    out_channels=64,
    num_outs=4,
    #trans_idx=-1,
)

loss_weights=dict(
        hm_weight=1,
        kps_weight=0.4,
        row_weight=1.,
        range_weight=1.,
    )

num_lane_classes=1
heads=dict(
    type='CondLaneHead',
    heads=dict(hm=num_lane_classes),
    in_channels=(64, ),
    num_classes=num_lane_classes,
    head_channels=64,
    head_layers=1,
    disable_coords=False,
    branch_in_channels=64,
    branch_channels=64,
    branch_out_channels=64,
    reg_branch_channels=64,
    branch_num_conv=1,
    hm_idx=2,
    mask_idx=0,
    compute_locations_pre=True,
    location_configs=dict(size=(batch_size, 1, 80, 200), device='cuda:0')
)

optimizer = dict(type='AdamW', lr=3e-4, betas=(0.9, 0.999), eps=1e-8)
optimizer = dict(type='SGD', lr=3e-3)

epochs = 40
total_iter = (88880 // batch_size) * epochs
total_iter = (3688 // batch_size) * epochs

import math
scheduler = dict(
    type = 'MultiStepLR',
    milestones=[15, 25, 35],
    gamma=0.1
)

seg_loss_weight = 1.0
eval_ep = 1
save_ep = 1 

img_norm = dict(
    mean=[75.3, 76.6, 77.6],
    std=[50.5, 53.8, 54.3]
)

img_height = 320 
img_width = 800
cut_height = 0 
ori_img_h = 590
ori_img_w = 1640

mask_down_scale = 4
hm_down_scale = 16
num_lane_classes = 1
line_width = 3
radius = 6
nms_thr = 4
img_scale = (800, 320)
crop_bbox = [0, 270, 1640, 590]
mask_size = (1, 80, 200)

train_process = [
    dict(type='Alaug',
    transforms=[dict(type='Compose', params=dict(bboxes=False, keypoints=True, masks=False)),
    dict(
        type='Crop',
        x_min=crop_bbox[0],
        x_max=crop_bbox[2],
        y_min=crop_bbox[1],
        y_max=crop_bbox[3],
        p=1),
    dict(type='Resize', height=img_scale[1], width=img_scale[0], p=1),
    dict(
        type='OneOf',
        transforms=[
            dict(
                type='RGBShift',
                r_shift_limit=10,
                g_shift_limit=10,
                b_shift_limit=10,
                p=1.0),
            dict(
                type='HueSaturationValue',
                hue_shift_limit=(-10, 10),
                sat_shift_limit=(-15, 15),
                val_shift_limit=(-10, 10),
                p=1.0),
        ],
        p=0.7),
    dict(type='JpegCompression', quality_lower=85, quality_upper=95, p=0.2),
    dict(
        type='OneOf',
        transforms=[
            dict(type='Blur', blur_limit=3, p=1.0),
            dict(type='MedianBlur', blur_limit=3, p=1.0)
        ],
        p=0.2),
    dict(type='RandomBrightness', limit=0.2, p=0.6),
    dict(
        type='ShiftScaleRotate',
        shift_limit=0.1,
        scale_limit=(-0.2, 0.2),
        rotate_limit=10,
        border_mode=0,
        p=0.6),
    dict(
        type='RandomResizedCrop',
        height=img_scale[1],
        width=img_scale[0],
        scale=(0.8, 1.2),
        ratio=(1.7, 2.7),
        p=0.6),
    dict(type='Resize', height=img_scale[1], width=img_scale[0], p=1),]
    ),
    dict(type='CollectLane',
        down_scale=mask_down_scale,
        hm_down_scale=hm_down_scale,
        max_mask_sample=5,
        line_width=line_width,
        radius=radius,
        keys=['img', 'gt_hm'],
        meta_keys=[
            'gt_masks', 'mask_shape', 'hm_shape',
            'down_scale', 'hm_down_scale', 'gt_points'
        ]
    ),
    #dict(type='Resize', size=(img_width, img_height)),
    dict(type='Normalize', img_norm=img_norm),
    dict(type='ToTensor', keys=['img', 'gt_hm'], collect_keys=['img_metas']),
]


val_process = [
    dict(type='Alaug',
        transforms=[dict(type='Compose', params=dict(bboxes=False, keypoints=True, masks=False)),
            dict(type='Crop',
            x_min=crop_bbox[0],
            x_max=crop_bbox[2],
            y_min=crop_bbox[1],
            y_max=crop_bbox[3],
            p=1),
        dict(type='Resize', height=img_scale[1], width=img_scale[0], p=1)]
    ),
    #dict(type='Resize', size=(img_width, img_height)),
    dict(type='Normalize', img_norm=img_norm),
    dict(type='ToTensor', keys=['img']),
]

# dataset_path = './data/CULane'
dataset_path = './data/Merge_data'
# val_path = './data/CULane'
dataset = dict(
    train=dict(
        type='CULane',
        data_root=dataset_path,
        split='train',
        processes=train_process,
    ),
    val=dict(
        type='CULane',
        data_root=dataset_path,
        split='test',
        processes=val_process,
    ),
    test=dict(
        type='CULane',
        data_root=dataset_path,
        split='test',
        processes=val_process,
    )
)


workers = 6
log_interval = 100
lr_update_by_epoch=True

Thank you so much

使用测试kaist数据集测试[CondLane]，最多只能检测出3条车道线，很明显的车道检测不出来，请问是限制了检测车道线数量了吗，在那里可以配置？ https://github.com/Turoad/lanedet/issues/58#issuecomment-1131143127 按照此处的配置方法似乎不管用。

Traceback (most recent call last): File "tools/detect.py", line 86, in process(args) File "tools/detect.py", line 68, in process cfg = Config.fromfile(args.config) File "d:\lanedet\lanedet\utils\config.py", line 180, in fromfile cfg_dict, cfg_text = Config._file2dict(filename) File "d:\lanedet\lanedet\utils\config.py", line 105, in _file2dict shutil.copyfile(filename, File "C:\Users\l00653465\Anaconda3\envs\lanedet\lib\shutil.py", line 264, in copyfile with open(src, 'rb') as fsrc, open(dst, 'wb') as fdst: PermissionError: [Errno 13] Permission denied: 'C:\Users\L00653~1\AppData\Local\Temp\tmphpklern8\tmpkydalnxp.py'

在进行训练和测试的时候都会报这个错

Hi Guys,

I am using this project on conda env with gpu configured. I was trying to just run the inference files first to try it out, but I get the following error:

Traceback (most recent call last): File "c:\CULane\lanedet\tools\detect.py", line 86, in process(args) File "c:\CULane\lanedet\tools\detect.py", line 72, in process detect = Detect(cfg) File "c:\CULane\lanedet\tools\detect.py", line 24, in init load_network(self.net, self.cfg.load_from) File "c:\culane\lanedet\lanedet\utils\net_utils.py", line 48, in load_network net.load_state_dict(pretrained_model['net'], strict=True) KeyError: 'net'

I have used following command: $ python detect.py' 'lanedet/configs/resa/resa34_culane.py' '--img' 'image\' '--load_from' 'C:\Users\blackbug\.cache\torch\hub\checkpoints\resnet34-333f7ec4.pth' '--savedir' './vis'

I tried to look at the model loaded from the downloaded resnet model file; it looks valid with all the trained layers, just "net" isnt part of the dictionary. Any help is appreciated! Thank you!