README
Introduction
Disclaimer: This dataset is organized and adapted from Phineas476/EmbSpatial-Bench. The original data was image format and has been converted here into a more accessible and easy-to-use format.
EmbSpatial-Bench is a benchmark for evaluating embodied spatial understanding of LVLMs. The benchmark is automatically derived from embodied scenes and covers 6 spatial relationships from an egocentric perspective. The constructed benchmark comprises a total of 3,640 QA pairs, covering 294 object categories and 6 spatial relationships.
Data Fields
| Field Name | Type | Description |
|---|---|---|
| data_source | string | Name or identifier of the data source |
| scene_id | string | Unique scene ID |
| question_id | string | Unique question ID |
| question | string | Text of the question |
| relation | string | Relationship described in the question |
| image | Image | Scene image data |
| answer_options | sequence | List of answer options |
| answer | int32 | Index of the correct answer |
| objects | sequence | List of objects in the image |
| bbox | sequence | Bounding box coordinates (x_min, y_min, x_max, y_max) |
| name | string | Name of the object |
| image_base64 | string | Image data in base64 encoding |
Contributors
3 commits
Similar repos
README
Introduction
Disclaimer: This dataset is organized and adapted from Phineas476/EmbSpatial-Bench. The original data was image format and has been converted here into a more accessible and easy-to-use format.
EmbSpatial-Bench is a benchmark for evaluating embodied spatial understanding of LVLMs. The benchmark is automatically derived from embodied scenes and covers 6 spatial relationships from an egocentric perspective. The constructed benchmark comprises a total of 3,640 QA pairs, covering 294 object categories and 6 spatial relationships.
Data Fields
| Field Name | Type | Description |
|---|---|---|
| data_source | string | Name or identifier of the data source |
| scene_id | string | Unique scene ID |
| question_id | string | Unique question ID |
| question | string | Text of the question |
| relation | string | Relationship described in the question |
| image | Image | Scene image data |
| answer_options | sequence | List of answer options |
| answer | int32 | Index of the correct answer |
| objects | sequence | List of objects in the image |
| bbox | sequence | Bounding box coordinates (x_min, y_min, x_max, y_max) |
| name | string | Name of the object |
| image_base64 | string | Image data in base64 encoding |
Similar repos
Linked in READMEs
Contributors
3 commits