Songyang Zhang

The code should work for this scenario as the broadcast will be conducted automatically. You can try this example ```bash >>> import torch >>> data = torch.randn(1,10, 512,512) >>> a...

First, in DisAlign, **all layers learned in stage-1 are fixed** during the second stage. Second, the proposed method mainly focuses on **long-tail**, which typically consists of many tail classes. Thus,...

Hi, the case described is imbalanced classification, not long-tail. long-tail means there exist **many** tail classes.(typically hundreds or thousands of classes in total)

@joishbader Hi, thanks for your work, do you have plan to release the code for IoU-Net.

@aromazyl https://www.cityscapes-dataset.com/

@YuwenXiong Hi, Thanks for your reply. I have implemented the evaluation on cityscapes. I would like to know the details of result from your paper(Table 4.). For Cityscapes(l=5), DFF 69.2...

@chowkamlee81 We have implemented a version of video semantic segmentation based on DFF https://github.com/tonysy/Deep-Feature-Flow-Segmentation

Could you share your env setting? PyTorch version? system version? - First, there exist two phases in the training phase described in the original paper. - Training stage - Finetune...

I use the multi-gpu(2-gpus), set batch size =32, get 77.3% for the first stage. I found the experiment results are random, performance range from(74.8 ~ 77.3). I plan to add...

I use the following code to plot 3D box ``` def plot_3d_bbox(img, label_info): alpha = label_info['alpha'] # theta_ray = label_info['theta_ray'] box_3d = [] center = label_info['location'] dims = label_info['dimension'] calib...