How to use a good trained model to predict multi angle photos—— Questions from new scholars

[graceforlife]

Thank you for sharing such beautiful code. I have a question. In readme, I can find the use of train, which is used to train models, and evaluate, which is used to evaluate models. But I went to predicate from there. For example, I had trained the model, and then I wanted to use three new kinds of campus photos to fuse the trained model into a 3D image. How should I operate, thank you!

[AlvinYH]

Hi, thanks for your interest in our work. Based on my understanding, your concern is about how to make predictions on your own dataset with the pre-trained model. To start with, we've modified the code and you can pull the recent release and refer to README.md. Specifically, it's required to implement the code to process your own dataset under the lib/dataset/ directory. You can refer to lib/dataset/shelf.py and rewrite the _get_db and _get_cam functions to take RGB images and camera params as input. Then modify the config file based on configs/shelf/jln64.yaml. Then you can adopt the same training strategy as what we use on the Campus and Shelf datasets, i.e. synthesizing training poses with Panoptic dataset.