cch2016

Results 8 comments of cch2016

About the robustness and portability of monocular 3D models

> You just need to make sure the calib matrix is formatted correctly, and the parameters can vary from camera to camera. We have done verification on the nuscenes data...

About the robustness and portability of monocular 3D models

@walzimmer You should use projection matrix from your own dataset. It's generalization is pretty good.

What time is the code available

Code! Code ! Code !

Can't get attribute 'CHJ_tiny_obj' on <module 'objloader'

> When running car_instance/demo.ipynb, I met the following error. Is there any clue to fix this? Many thanks! > > ----> 8 visualizer.load_car_models() > 9 image_vis, mask, depth = visualizer.showAnn(setting.image_name)...

pytorch version of DAN loss : I dont't find betals param of multi guassian_kernels

I just see "sum" , mean just "sum" and dont't have "beta " parameter

In the DAN Pytorch implementation, Code about Bata for different kernels?

I have also find this problems

The problem is “The model and loaded state dict do not match exactly”

I replace backbone resnet50 with resnext101, I meet `missing keys in source state_dict ` problem: ``` 2020-08-11 10:25:03,000 - mmdet - INFO - load model from: open-mmlab://resnext101_32x4d 2020-08-11 10:25:27,770 -...

How can I determine which region of the image the model is focusing on when answering a specific question?

> Thank you for your interest in our work. You can try to achieve such effect by adding an extra prompt `Answer the question according to the region[[x1, y1, x2,...