# Twitter100k: A Real-world Dataset for Weakly Supervised Cross-Media Retrieval

arXiv: https://arxiv.org/abs/1703.06618 · HTML: https://mlsanity.com/paper/1703.06618/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 20 Mar 2017.

## Built on by (most cited first, 7 of 7)

- [Product1M: Towards Weakly Supervised Instance-Level Product Retrieval via Cross-modal Pretraining](https://mlsanity.com/paper/2107.14572/) (2021)
- [RpBERT: A Text-image Relation Propagation-based BERT Model for Multimodal NER](https://mlsanity.com/paper/2102.02967/) (2021)
- [M5Product: Self-harmonized Contrastive Learning for E-commercial Multi-modal Pretraining](https://mlsanity.com/paper/2109.04275/) (2021)
- [Towards Continual Egocentric Activity Recognition: A Multi-modal Egocentric Activity Dataset for Continual Learning](https://mlsanity.com/paper/2301.10931/) (2023)
- [MIKO: Multimodal Intention Knowledge Distillation from Large Language Models for Social-Media Commonsense Discovery](https://mlsanity.com/paper/2402.18169/) (2024)
- [Entity-Graph Enhanced Cross-Modal Pretraining for Instance-level Product Retrieval](https://mlsanity.com/paper/2206.08842/) (2022)
- [The Role of the Input in Natural Language Video Description](https://mlsanity.com/paper/2102.05067/) (2021)

## Abstract (arXiv)

This paper contributes a new large-scale dataset for weakly supervised cross-media retrieval, named Twitter100k. Current datasets, such as Wikipedia, NUS Wide and Flickr30k, have two major limitations. First, these datasets are lacking in content diversity, i.e., only some pre-defined classes are covered. Second, texts in these datasets are written in well-organized language, leading to inconsistency with realistic applications. To overcome these drawbacks, the proposed Twitter100k dataset is characterized by two aspects: 1) it has 100,000 image-text pairs randomly crawled from Twitter and thus has no constraint in the image categories; 2) text in Twitter100k is written in informal language by the users. Since strongly supervised methods leverage the class labels that may be missing in practice, this paper focuses on weakly supervised learning for cross-media retrieval, in which only text-image pairs are exploited during training. We extensively benchmark the performance of four subspace learning methods and three variants of the Correspondence AutoEncoder, along with various text features on Wikipedia, Flickr30k and Twitter100k. Novel insights are provided. As a minor contribution, inspired by the characteristic of Twitter100k, we propose an OCR-based cross-media retrieval method. In experiment, we show that the proposed OCR-based method improves the baseline performance.