53 datasets found

The SubICap-1k dataset is a smaller version of the MS-COCO dataset, containing 1,000 images with 1,000 captions each.

The MSCOCO caption challenge dataset is a subset of the MSCOCO caption dataset, containing 113,287 training images, 5,000 validation images, and 5,000 test images.

The MSCOCO caption dataset is a large-scale image captioning dataset. It consists of 123,000 images with 5 captions each.

Visual language grounding is widely studied in modern neural networks, which typically adopts an encoder-decoder framework consisting of a convolutional neural network (CNN) for...

A dataset of images annotated with captions and semantic tuples, created by training a model to predict semantic tuples from image captions.

The dataset used in the paper is Flickr30K and MSCOCO, which are used for image-text matching and image captioning tasks.

The Flickr 30k dataset is a large-scale image captioning dataset containing 30,000 images with 30 captions each.

The MSCOCO 2014 captions dataset contains 123,293 images, split into a 82,783 image training set and a 40,504 image validation set. Each image is labeled with five...

The High Quality Image Text Pairs (HQITP-134M) dataset consists of 134 million diverse and high-quality images paired with descriptive captions and titles.

The dataset used for training the Brazilian Portuguese version of the GRIT model, a translation of the COCO dataset.

The dataset used in this paper for semantic communication, consisting of images and their corresponding captions.

The Pascal Flickr dataset is a collection of captions for images from Flickr.

Flickr8K dataset is a collection of 8,000 images with 5 sentences describing each image content

The dataset used in the paper is a image captioning task.

Image captioning is a fundamental task in vision-language understanding, where the model predicts a textual informative caption to a given input image.

The dataset used in the paper to evaluate the effectiveness of context-driven entity-aware captioning.

The Winoground dataset consists of 400 items, each containing two image-caption pairs (I0, C0), (I1, C1).

CC12M dataset is used for training and testing the proposed method. It contains 12 million images with 12 million captions.

Flickr8K-Expert dataset is used for evaluating the proposed method. It contains 8,000 images with 8,000 captions.

The Composite dataset, containing 11,985 human judgments over Flickr 8K, Flickr 30K, and COCO captions.