SpadeLiu
SpadeLiu
> @SpadeLiu After I trained Baseline8 epochs by myself, I tested grafting VGG features with its aggregation module, but the results seemed to be unstable. Do you have the same...
Hi, maybe the clean pass version, I checked the name of my files. There are 35454 training pairs in total. ---Original--- From: ***@***.***> Date: Mon, Jan 17, 2022 21:50 PM...
I am sure the dataset is downloaded from this website, but I can not confirm whether clean pass or final pass, I think clean pass. Maybe you can check the...
> thanks for your good work, but i have downloaded the sceneflow weights, and trained the kitti_15 datasets using the kitti_ft.py as the command "python KITTI_ft.py --data_path /data/kitti_2015/training/ --gpu_id 4...
I didn't receive an email, please check again.
> Here is the address: [email protected]
@luyao-cv Please check your email.
> We split the 200 pairs into a training set including 160 pairs and a validation set including 40 pairs. Could you please provide the name list of the training...
Hi, The model used in this paper is not so robust, if there is a large domain gap between KITTI and RobotCar, the model can not generalize well. You can...
Here are some examples: "Monkaa": /media/SceneFlow/monkaa_frames_cleanpass/a_rain_of_stones_x2/left/0000.png "Driving": /media/SceneFlow/driving_frames_cleanpass/15mm_focallength/scene_backwards/fast/left/0001.png "FlyingThings3D": /media/ScemeFlow/flying_frames_cleanpass/TRAIN/A/0000/left/0006.png