ORGANISATION

TamilCOCO

TamilCOCO is a novel bilingual image captioning dataset specifically designed for Tamil, a low-resource language. This dataset facilitates research in image captioning, cross-lingual natural language processing, and culturally adapted AI applications.

About Dataset

TamilCOCO is a bilingual vision-language dataset developed to support image captioning and multimodal artificial intelligence research for the Tamil language. The dataset combines images from the widely used MS COCO benchmark with Tamil-language captions, creating a valuable resource for connecting visual content with natural language descriptions in an Indic language. By extending image captioning resources to Tamil, the dataset helps address the scarcity of multimodal datasets for Indian languages and contributes to the growth of inclusive AI technologies and multilingual language resources. The citation for the article - Prakash, J. V., Vijay, S. A. A., Balamurugan, R., Sudharsan, S., Sanjai, T., & Sameer, A. A. (2026). TamilCOCO: A culturally-enriched Tamil image captioning dataset derived from MSCOCO. Data and Information Management, 10(3), 100137. https://doi.org/10.1016/j.dim.2026.100137. This dataset was identified and facilitated for onboarding as part of the Dataset Onboarding Support Team (DOST) initiative led by CivicDataLab (CDL), partnering with the Gates Foundation in collaboration with BHASHINI. CivicDataLab provided support for dataset discovery, metadata preparation and onboarding facilitation. All dataset ownership and intellectual property rights remain with the original author(s).

Purpose of Dataset

The Purpose Of This Dataset Is To Support The Development And Evaluation Of Image Captioning Systems For The Tamil Language. It Enables Research On Generating Natural Language Descriptions From Visual Content, Understanding Relationships Between Images And Text, And Building Multimodal Machine Learning Models That Integrate Computer Vision And Natural Language Processing. The Dataset Can Be Used For Image Caption Generation, Image-text Retrieval, Visual Question Answering, Cross-lingual Vision-language Research, And The Advancement Of Tamil Language Resources In Multimodal Ai.

Dataset Metadata

License

MIT

Geographical coverage

India

Sector

Education and Skill Development

Author

Jothi Prakash V, Arul Antran Vijay S, Balamurugan R, Sudharsan S, Sanjai T and Ahamed Sameer A

Source Organisation

Digital India BHASHINI Division

Uploaded by

Nikil Augustine

Data Quality Score (Beta)

Dataset type

Structured

Frequency

Static

Time Granularity

Static

Year range

09/01/2025 - 02/02/2025

Date & Time

10/07/26 10:48:08

Visibility

Open

Primary Key / Indicator

Image_id

Hosted / Redirected

Redirected

Data Type

Hybrid

If Redirection which source

Https://github.com/jothiprakashv/tamilcoco

Data Collection Method

Tamilcoco Uses A Semi-automated Annotation Framework That Integrates Multilingual Models, A Cultural Adaptation Module (Cam), And Iterative Community Validation, Ensuring Both Semantic Accuracy And Cultural Fidelity. The Dataset Was Evaluated Using Metrics—bleu (0.68), Meteor (0.83), Cider (1.48), And Spice (0.53)—alongside A Cultural Relevance Score (Crs) Averaging 0.86, Highlighting Its Cultural Alignment.