metadata

license: cc-by-nc-4.0
language:
  - en
pretty_name: Stark
tags:
  - multi-modal dialogue
annotation_creators:
  - machine-generated
task_ids:
  - conversational
dataset_size: 899K

Dataset Card for Stark

🏠 Homepage | 💻 Github | 📄 Arxiv | 📕 PDF

List of Provided Model Series

Ultron-Summarizer-Series: 🤖 Ultron-Summarizer-1B | 🤖 Ultron-Summarizer-3B | 🤖 Ultron-Summarizer-8B
Ultron 7B: 🤖 Ultron-7B

🚨 Disclaimer: All models and datasets are intended for research purposes only.

Dataset Description

Repository: Code
Paper: Stark: Social Long-Term Multi-Modal Conversation with Persona Commonsense Knowledge
Point of Contact: Young-Jun Lee

Dataset Summary

Stark is a publicly available, large-scale, long-term multi-modal conversation dataset that encompasses a diverse range of social personas, multi-modality formats, time intervals, and images. To automatically construct Stark, we introduce a novel multi-modal contextualization framework, MCU, which generates long-term multi-modal dialogues distilled from ChatGPT and our proposed Plan-and-Execute Image Aligner. An overview of MCU and an example from Stark are illustrated below.

Note: The Stark dataset contains many images that originate from different sources, including a personalized text-to-image generator (i.e., PhotoMaker), Bing Search, and Image DB retrieval, thanks to our proposed Plan-and-Execute Image Aligner to ensure the high quality of the Stark dataset. Therefore, we divide the entire collection of images into two repositories based on the source:

This Repository: Contains images sourced from CC12M and RedCaps12M, which provide the image URL links.
Another Repository: Contains images from the personalized text-to-image generator, Bing Search, and Image DB retrieval (excluding CC12M and RedCaps12M). This dataset is provided in this repository.

⚠️ Warning: Since the images provided in this repository are sourced from CC12M and RedCaps12M, some images may not be accessible if the original URL owners have restricted access or removed them.

For image database retrieval, we use CLIP (i.e., ViT-L/14@336px) to retrieve images from the prepared source image datasets: CC12M, RedCaps12M, ChartQA, AI2D, and MathVision. For efficient searching, we utilize AutoFaiss.

Dataset Structure

Since the number of images is large (roughly 899K), we store and provide the image datasets in WebDataset format for efficiency.

Field	Type	Description
`key`	str	A unique identifier for each data entry in the dataset.
`url`	str	The URL path to the image stored in the dataset repository on HuggingFace. All URLs point to the base HuggingFace repository where images are stored.
`json`	dict	Contains additional metadata and information for each image, structured as a dictionary. The JSON field typically includes the following keys: `image_source`, `image_url`, `index`.

image_source: Source of the image (cc12m, redcaps12m), where:
- "cc12m": Image from CC12M via Image DB retrieval.
- "redcaps12m": Image from the RedCaps12M dataset via Image DB retrieval.
image_url: External URL where the image was originally sourced.
index: A unique index identifier for each image within the dataset, which is the same as the key field.

Languages

English

Further Details and Limitations

For additional information and limitations, please refer to our paper.

License and Recommendations

The Stark dataset is intended for research purposes only. Despite our efforts to generate high-quality and diverse personalized images, users should be mindful of ethical considerations when utilizing the dataset.

Acknowledgements

This work was supported by a grant from the KAIST-KT joint research project through AI Tech Lab, Institute of Convergence Technology, funded by KT [Project No. G01230605, Development of Task-oriented Persona-based Dialogue Generation Combining Multi-modal Interaction and Knowledge Modeling].

Citation

If you find the resources in this repository useful, please cite our work: