AnyNet
AnyNet copied to clipboard
mileyan
(ICRA) Anytime Stereo Image Depth Estimation on Mobile Devices
Hi @mileyan, Thanks for sharing the amazing work. I am curious about the `_build_volume_2d3` function in line 118 of anynet.py. https://github.com/mileyan/AnyNet/blob/83badc75af596e76edce67e4917ba4fbacbb7c1d/models/anynet.py#L118 Why `batch_disp = batch_disp - batch_shift.float()` and not `batch_disp...
Environment: GTX 3090 torch1.7.0+cu110 torchvision0.8.1 python3.8 File: if __name__ == '__main__': # main() img_l = torch.randn(1,3,512,256).cuda() img_r = torch.randn(1,3,512,256).cuda() f = r"../results/pretrained_anynet/checkpoint.tar" checkpoint = torch.load(f) checkpoint['state_dict'] = proc_nodes_module(checkpoint, 'state_dict') model...
Hi @mileyan , I am implementing AnyNet for my project. However I have following questions: 1) I have not been able to compile SPN Module. It gives me the error...
I downloaded your pre-trained model and used the following command to test. python finetune.py --maxdisp 192 --datapath path-to-kitti2015/training/ \ --save_path results/kitti2015 --datatype 2015 --pretrained checkpoint/kitti2015_ck/checkpoint.tar \ --split_file checkpoint/kitti2015_ck/split.txt --evaluate without...
hello, Thanks for the great work , I want to ask that is it estimating depth of 2 images (pair) or estimating a real time depth from two cameras? Please...
Hi, First of all thanks for your great efforts on this research. I have a question during the evaluation process of AnyNet through Jetson TX2. After training AnyNet with GTX...
Hi, thanks for your great work! What backend did you use to run inference on the device? Have you used PyTorch model as is or converted it to TensorRT?
First of all I congratulate you on the result of the project I'm interested in testing a tensorflow lite version of this model. Do you have any?
we want to train AnyNet to calculate the loss on depth not disparity as it will be more efficient as discussed in Pseudo Lidar to use AnyNet with its high...
