zhyever

I will release configs for BinsFormer on KITTI soon, and I would like to release all the trained models and logs. Thanks for your attention.

Unfortunately, I did not consider the FP16 training when I implemented this repo, so the interfaces in codes are not accomplished now. But I think you can follow the docs...

Following previous works, pixels at these places are masked during evaluation. In other words, they do not take counts in evaluation.

I have implemented several codes to convert the model to ONNX but not release in this repo now. The most tricky issue is whether the ONNX has supported ops in...

Sorry for the late reply, I missed your questions. Yes, the input of the model during converting should be single. In other words, the forward function should be like: `def...

Same issue. I think it leads to shifts in visualizations of SMOKE predictions on images.

> Same issue. I think it leads to shifts in visualizations of SMOKE predictions on images. Sorry, the shift of gt is caused by unchanged cam intrinsic param in rescaling...

Thanks for your appreciation. I think you have already had a gooood start while there are several small bugs. As you can refer [here](https://github.com/zhyever/Monocular-Depth-Estimation-Toolbox/blob/9e490dda2940480cbedbf2893065b1e040691677/depth/models/depther/base.py#L111), we need pass 'img' and 'img_metas'...

Thanks a lot for your contributions. I will consider it carefully and enhance this toolbox! To be honest, it is interesting to visualize the depth results, which makes the model...

@Chun-wei-Ho Thanks a lot for your support. I think all the notifications of adopting a custom dataset contain in your answer. @Boraly That's an error about incorrectly loading the GT...