כתוביות_רעיוניות

הפניות:

השתמש בפקודה הבאה כדי לטעון מערך נתונים זה ב-TFDS:

ds = tfds.load('huggingface:conceptual_captions')
  • תיאור :
Image captioning dataset
The resulting dataset (version 1.1) has been split into Training, Validation, and Test splits. The Training split consists of 3,318,333 image-URL/caption pairs, with a total number of 51,201 total token types in the captions (i.e., total vocabulary). The average number of tokens per captions is 10.3 (standard deviation of 4.5), while the median is 9.0 tokens per caption. The Validation split consists of 15,840 image-URL/caption pairs, with similar statistics.
  • רישיון : אין רישיון ידוע
  • גרסה : 1.1.0
  • פיצולים :
לְפַצֵל דוגמאות
'train' 3318333
'validation' 15840
  • מאפיינים :
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "caption": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "url": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

ללא תווית

השתמש בפקודה הבאה כדי לטעון מערך נתונים זה ב-TFDS:

ds = tfds.load('huggingface:conceptual_captions/unlabeled')
  • תיאור :
Google's Conceptual Captions dataset has more than 3 million images, paired with natural-language captions.
In contrast with the curated style of the MS-COCO images, Conceptual Captions images and their raw descriptions are harvested from the web,
and therefore represent a wider variety of styles. The raw descriptions are harvested from the Alt-text HTML attribute associated with web images.
The authors developed an automatic pipeline that extracts, filters, and transforms candidate image/caption pairs, with the goal of achieving a balance of cleanliness,
informativeness, fluency, and learnability of the resulting captions.
  • רישיון : ניתן להשתמש במערך הנתונים באופן חופשי לכל מטרה, אם כי הודאה של Google LLC ("Google") כמקור הנתונים תתקבל בברכה. מערך הנתונים מסופק "כמות שהוא" ללא כל אחריות, מפורשת או משתמעת. Google מתנערת מכל אחריות לכל נזק, ישיר או עקיף, הנובע מהשימוש במערך הנתונים.

  • גרסה : 0.0.0

  • פיצולים :

לְפַצֵל דוגמאות
'train' 3318333
'validation' 15840
  • מאפיינים :
{
    "image_url": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "caption": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

מסומן

השתמש בפקודה הבאה כדי לטעון מערך נתונים זה ב-TFDS:

ds = tfds.load('huggingface:conceptual_captions/labeled')
  • תיאור :
Google's Conceptual Captions dataset has more than 3 million images, paired with natural-language captions.
In contrast with the curated style of the MS-COCO images, Conceptual Captions images and their raw descriptions are harvested from the web,
and therefore represent a wider variety of styles. The raw descriptions are harvested from the Alt-text HTML attribute associated with web images.
The authors developed an automatic pipeline that extracts, filters, and transforms candidate image/caption pairs, with the goal of achieving a balance of cleanliness,
informativeness, fluency, and learnability of the resulting captions.
  • רישיון : ניתן להשתמש במערך הנתונים באופן חופשי לכל מטרה, אם כי הודאה של Google LLC ("Google") כמקור הנתונים תתקבל בברכה. מערך הנתונים מסופק "כמות שהוא" ללא כל אחריות, מפורשת או משתמעת. Google מתנערת מכל אחריות לכל נזק, ישיר או עקיף, הנובע מהשימוש במערך הנתונים.

  • גרסה : 0.0.0

  • פיצולים :

לְפַצֵל דוגמאות
'train' 2007090
  • מאפיינים :
{
    "image_url": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "caption": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "labels": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "MIDs": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "confidence_scores": {
        "feature": {
            "dtype": "float64",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}