The structure of the dataset

naver-ai / egtr

[CVPR 2024 Best paper award candidate] EGTR: Extracting Graph from Transformer for Scene Graph Generation

https://arxiv.org/abs/2404.02072

Apache License 2.0

38 stars 0 forks source link

The structure of the dataset #3

Open slcheng97 opened 1 week ago

slcheng97 commented 1 week ago

Could you show how the stureture of the dataset should be? When I run your code, I alwayes encounter data loading problem as below:

jinbae commented 1 week ago

You can see the structure of the dataset in the README as follows:

dataset
│   
└───visual_genome
│   │
│   └───images
│       train.json
│       val.json
│       test.json
│       rel.json
│   
└───open-imagev6
    │
    └───images
        annotations

Please note that all 100K images are in dataset/visual_genome/images.