YOLOP
YOLOP copied to clipboard
hustvl
You Only Look Once for Panopitic Driving Perception.（MIR2022）
Hello the dataset of bdd 100k according to one json file?    I have changed the structure of datasets like your files but i couldn't find the correct...
如题，训练的时候，GPU利用率忽高忽低，导致模型训练很慢，时间很长。尝试过以下的措施： dataloader的时候将 pin_mermory改为True， num_workers 改为4， 8， 16 都尝试了，都没有得到改善。 有谁知道怎么解决吗？
I set the `default.py` like below: ``` # Alternating optimization _C.TRAIN.SEG_ONLY = False # Only train two segmentation branchs _C.TRAIN.DET_ONLY = False # Only train detection branch _C.TRAIN.ENC_SEG_ONLY = False...
nice work！ What kind of tasks are suitable for training with a multitasking framework? Suppose that the data of different tasks are different, such as the two tasks of vehicle...
Hello, may I ask what the specifications and model of your hardware are that can achieve 4X more Fps?
I would like to ask about the problem related to training process. I put my dataset as below   However, I had a problem as below **self = reduction.pickle.load(from_parent)...
May I ask if YOLOP wants to TRAIN the single task separately, except changing _C. train.drivable_only or _C. train.lane_only or _C. train.det_only to True in defaut.py, what else needs to...
I wanted to fine-tune only the detection head of the model, so I set _C.TRAIN.DET_ONLY=True in the config file. However, I noticed that the encoder is also getting updated, and...
Hi I want to run this repo on my RTX 4070. I do not have ZED Camera. I do not want to use it. Is there any way to run...
