10000 Evaluation on DTU, Re10k and Mipnerf · Issue #5 · MegaScenes/nvs · GitHub
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content
Evaluation on DTU, Re10k and Mipnerf #5
Open
@D-Nilsson

Description

@D-Nilsson

Hi, thanks for the excellent work! I am interested in evaluating the trained model on DTU, Re10k and Mipnerf to reproduce Table 2 in the paper. There is some code in test.py (copied below) and a flag --dataset, but these data loading classes are not defined or imported in the released code. I would be grateful if this was supported.

    if dataset_name == 'dtu':
        dataset = DTUDataset(pose_cond=train_configs['pose_cond'])
    elif dataset_name == 're10k':
        dataset = Re10kDataset(pose_cond=train_configs['pose_cond'])
    elif dataset_name == 'mipnerf':
        dataset = MipnerfDataset(pose_cond=train_configs['pose_cond'])
    elif dataset_name == 'megascenes':
        dataset = PairedDataset(pose_cond=train_configs['pose_cond'], split='test')

Thanks!

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions

      0