MagicDrive
MagicDrive copied to clipboard
cure-lab
[ICLR24] Official implementation of the paper “MagicDrive: Street View Generation with Diverse 3D Geometry Control”
Hello, after I prepare the cache files in h5 format, how should I call these cache files during training?
Hello, I retrain and get the result as your paper of the AP、NDS and FID at 224x400 resolution, but failed using 272x736 resolution. Can your give me more detail about...
Thanks for your great work! I have a question about your implementation: - In Figure 3, the embeded road map is interacted with Seq Embedding and then added to input...
Hi, I assume the stable diff model you used is deprecated since the [link ](https://huggingface.co/runwayml/stable-diffusion-v1-5) doesn't work. Which huggingface model should be used instead? Or do you have the folders...
I have a question regarding Table 6 in the paper of the magicDrive v7 version. If the training period is adopted at 2x, the difference in the cam-only results between...
Hi, I was wondering how many epochs you trained on your released checkpoint? I noticed that the fid of the pre-generated samples you provided for download in the "To test...
Hi, I would like to know how you calculate the FVD. Because the selection of perception models will greatly affect the FVD. In order to make a fair comparison, I...
Weather
 If I want to generate a rainy day scene during the testing process, where should I modify the code？ Thanks！
In image generation, different views use the same noise as validation initial. `torch.stack([latents]*6,dim=0)` But in the video generation, it seems not work to do the same way. The nosie between...
