compguesswhat

参考:

compguesswhat-original

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:compguesswhat/compguesswhat-original')
  • 说明
CompGuessWhat?! is an instance of a multi-task framework for evaluating the quality of learned neural representations,
        in particular concerning attribute grounding. Use this dataset if you want to use the set of games whose reference
        scene is an image in VisualGenome. Visit the website for more details: https://compguesswhat.github.io
  • 许可:无已知许可
  • 版本:0.2.0
  • 拆分
拆分 样本
'test' 9621
'train' 46341
'validation' 9738
  • 特征
{
    "id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "target_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "timestamp": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "status": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "image": {
        "id": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "file_name": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "flickr_url": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "coco_url": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "height": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "width": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "visual_genome": {
            "width": {
                "dtype": "int32",
                "id": null,
                "_type": "Value"
            },
            "height": {
                "dtype": "int32",
                "id": null,
                "_type": "Value"
            },
            "url": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "coco_id": {
                "dtype": "int32",
                "id": null,
                "_type": "Value"
            },
            "flickr_id": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "image_id": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            }
        }
    },
    "qas": {
        "feature": {
            "question": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "answer": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "id": {
                "dtype": "int32",
                "id": null,
                "_type": "Value"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "objects": {
        "feature": {
            "id": {
                "dtype": "int32",
                "id": null,
                "_type": "Value"
            },
            "bbox": {
                "feature": {
                    "dtype": "float32",
                    "id": null,
                    "_type": "Value"
                },
                "length": 4,
                "id": null,
                "_type": "Sequence"
            },
            "category": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "area": {
                "dtype": "float32",
                "id": null,
                "_type": "Value"
            },
            "category_id": {
                "dtype": "int32",
                "id": null,
                "_type": "Value"
            },
            "segment": {
                "feature": {
                    "feature": {
                        "dtype": "float32",
                        "id": null,
                        "_type": "Value"
                    },
                    "length": -1,
                    "id": null,
                    "_type": "Sequence"
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

compguesswhat-zero_shot

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:compguesswhat/compguesswhat-zero_shot')
  • 说明
CompGuessWhat?! is an instance of a multi-task framework for evaluating the quality of learned neural representations,
        in particular concerning attribute grounding. Use this dataset if you want to use the set of games whose reference
        scene is an image in VisualGenome. Visit the website for more details: https://compguesswhat.github.io
  • 许可:无已知许可
  • 版本:0.2.0
  • 拆分
拆分 样本
'nd_test' 13836
'nd_valid' 5343
'od_test' 13300
'od_valid' 5372
  • 特征
{
    "id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "target_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "status": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "image": {
        "id": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "file_name": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "coco_url": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "height": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "width": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "license": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "open_images_id": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "date_captured": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        }
    },
    "objects": {
        "feature": {
            "id": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "bbox": {
                "feature": {
                    "dtype": "float32",
                    "id": null,
                    "_type": "Value"
                },
                "length": 4,
                "id": null,
                "_type": "Sequence"
            },
            "category": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "area": {
                "dtype": "float32",
                "id": null,
                "_type": "Value"
            },
            "category_id": {
                "dtype": "int32",
                "id": null,
                "_type": "Value"
            },
            "IsOccluded": {
                "dtype": "int32",
                "id": null,
                "_type": "Value"
            },
            "IsTruncated": {
                "dtype": "int32",
                "id": null,
                "_type": "Value"
            },
            "segment": {
                "feature": {
                    "MaskPath": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "LabelName": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "BoxID": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "BoxXMin": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "BoxXMax": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "BoxYMin": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "BoxYMax": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "PredictedIoU": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "Clicks": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}