How to detect multiple objects using YOLOP?

[PigLogic-Cyber]

First of all, thx for your great jobs! The YOLOP model right now seems to have the ability to detect only the car, if I would like to detect more classes of object, what parameters should I modified? I had already tried to modify model.nc in train.py to be 41, change the single_cls to be False in bdd.py, uncomment the bdd_labels dict in convert.py, but still I got the error said:

Traceback (most recent call last): File "tools/train.py", line 406, in main() File "tools/train.py", line 333, in main train(cfg, train_loader, model, criterion, optimizer, scaler, File "/home/roy/Github/YOLOP/lib/core/function.py", line 77, in train total_loss, head_losses = criterion(outputs, target, shapes,model) File "/home/roy/anaconda3/envs/pytorch/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl return forward_call(*input, **kwargs) File "/home/roy/Github/YOLOP/lib/core/loss.py", line 50, in forward total_loss, head_losses = self._forward_impl(head_fields, head_targets, shapes, model) File "/home/roy/Github/YOLOP/lib/core/loss.py", line 96, in _forward_impl iou = bbox_iou(pbox.T, tbox[i], x1y1x2y2=False, CIoU=True) # iou(prediction, target) File "/home/roy/Github/YOLOP/lib/core/general.py", line 38, in bbox_iou print(box1[0] - box1[2] / 2) File "/home/roy/anaconda3/envs/pytorch/lib/python3.8/site-packages/torch/_tensor.py", line 249, in repr return torch._tensor_str._str(self) File "/home/roy/anaconda3/envs/pytorch/lib/python3.8/site-packages/torch/_tensor_str.py", line 415, in _str return _str_intern(self) File "/home/roy/anaconda3/envs/pytorch/lib/python3.8/site-packages/torch/_tensor_str.py", line 390, in _str_intern tensor_str = _tensor_str(self, indent) File "/home/roy/anaconda3/envs/pytorch/lib/python3.8/site-packages/torch/_tensor_str.py", line 251, in _tensor_str formatter = _Formatter(get_summarized_data(self) if summarize else self) File "/home/roy/anaconda3/envs/pytorch/lib/python3.8/site-packages/torch/_tensor_str.py", line 90, in init nonzero_finite_vals = torch.masked_select(tensor_view, torch.isfinite(tensor_view) & tensor_view.ne(0)) RuntimeError: CUDA error: device-side assert triggered CUDA kernel errors might be asynchronously reported at some other API call,so the stacktrace below might be incorrect. For debugging consider passing CUDA_LAUNCH_BLOCKING=1.

Please help me out!

[xinshoushangludd]

hello,I'm doing the same research. Have you ever solved the problem?
please help me.

[PigLogic-Cyber]

hello,I'm doing the same research. Have you ever solved the problem? please help me.

Try to modify the dict parameters in convert.py file, works for me, good luck, have fun : - )

[Nanjiangang]

Hello, what do you mean by “Try to modify the dict parameters in convert.py file”？ Can you tell us more about it？Thanks！

[Nanjiangang]

I set up ： model.nc=41 in train.py, single_cls = False in bdd.py， id_dict = {'xx':0, 'xx': 1,…… 'xx': 40} in convert.py but still I got the error said: Traceback (most recent call last): File "tools/train.py", line 395, in main() File "tools/train.py", line 323, in main epoch, num_batch, num_warmup, writer_dict, logger, device, rank) File "F:\yoloppp\YOLOP-main\lib\core\function.py", line 77, in train total_loss, head_losses = criterion(outputs, target, shapes,model) File "D:\anaconda\envs\yolop\lib\site-packages\torch\nn\modules\module.py", line 889, in _call_impl result = self.forward(*input, **kwargs) File "F:\yoloppp\YOLOP-main\lib\core\loss.py", line 50, in forward total_loss, head_losses = self._forward_impl(head_fields, head_targets, shapes, model) File "F:\yoloppp\YOLOP-main\lib\core\loss.py", line 107, in _forward_impl t[range(n), tcls[i]] = cp IndexError: index 26 is out of bounds for dimension 1 with size 1

[hetao828]

这个是因为YOLOP.py搭建网络里面nc默认设置成了1，需要改为我们自己的nc（例如41）

[hetao828]

like this!

[wangshichang-726]

按照之前的步骤操作，可以训练，但是将要训练下一轮的时候出现如下错误： Epoch: [1][1340/1342] Time 396.972s (200.944s) Speed 0.0 samples/s Data 396.692s (200.649s) Loss 0.69399 (0.75016) 0%| | 0/146 [00:00<?, ?it/s]F:\Anaconda\envs\yoloair\lib\site-packages\torch\nn\functional.py:3060: UserWarning: Default upsampling behavior when mode=bilinear is changed to align_corners=False since 0.4.0. Please specify align_corners=True if the old behavior is desired. See the documentation of nn.Upsample for details. warnings.warn("Default upsampling behavior when mode={} is changed " 0%| | 0/146 [00:01<?, ?it/s] Traceback (most recent call last): File "tools/train.py", line 395, in main() File "tools/train.py", line 330, in main da_segment_results,ll_segment_results,detect_results, total_loss,maps, times = validate( File "D:\YOLOP\YOLOP-main\lib\core\function.py", line 301, in validate label_det_pred = f'{names[int(cls)]} {conf:.2f}' KeyError: 35

[Lucas-TY]

按照之前的步骤操作，可以训练，但是将要训练下一轮的时候出现如下错误： Epoch: [1][1340/1342] Time 396.972s (200.944s) Speed 0.0 samples/s Data 396.692s (200.649s) Loss 0.69399 (0.75016) 0%| | 0/146 [00:00<?, ?it/s]F:\Anaconda\envs\yoloair\lib\site-packages\torch\nn\functional.py:3060: UserWarning: Default upsampling behavior when mode=bilinear is changed to align_corners=False since 0.4.0. Please specify align_corners=True if the old behavior is desired. See the documentation of nn.Upsample for details. warnings.warn("Default upsampling behavior when mode={} is changed " 0%| | 0/146 [00:01<?, ?it/s] Traceback (most recent call last): File "tools/train.py", line 395, in main() File "tools/train.py", line 330, in main da_segment_results,ll_segment_results,detect_results, total_loss,maps, times = validate( File "D:\YOLOP\YOLOP-main\lib\core\function.py", line 301, in validate label_det_pred = f'{names[int(cls)]} {conf:.2f}' KeyError: 35

search for nc = 1 change them to your class number

[L1NINE]

could you tell me more detail about it plz？I have no idea on where I can find nc=1 except model.nc in train.py...

[RahulRewale]

I created a fork and updated the code for BDD's 13 classes. Check it out here.