# A Strong and Robust Baseline for Text-Image Matching

arXiv: https://arxiv.org/abs/1906.01205 · HTML: https://mlsanity.com/paper/1906.01205/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 4 Jun 2019.

## Built on by (most cited first, 9 of 9)

- [TS2-Net: Token Shift and Selection Transformer for Text-Video Retrieval](https://mlsanity.com/paper/2207.07852/) (2022)
- [Cross Modal Retrieval with Querybank Normalisation](https://mlsanity.com/paper/2112.12777/) (2021)
- [Balance Act: Mitigating Hubness in Cross-Modal Retrieval with Query and Gallery Banks](https://mlsanity.com/paper/2310.11612/) (2023)
- [Dual-Modal Attention-Enhanced Text-Video Retrieval with Triplet Partial Margin Contrastive Learning](https://mlsanity.com/paper/2309.11082/) (2023)
- [NeighborRetr: Balancing Hub Centrality in Cross-Modal Retrieval](https://mlsanity.com/paper/2503.10526/) (2025)
- [Intra-Modal Constraint Loss For Image-Text Retrieval](https://mlsanity.com/paper/2207.05024/) (2022)
- [HNC: Leveraging Hard Negative Captions towards Models with Fine-Grained Visual-Linguistic Comprehension Capabilities](https://mlsanity.com/paper/2605.06157/) (2026)
- [DEMO: A Statistical Perspective for Efficient Image-Text Matching](https://mlsanity.com/paper/2405.11496/) (2024)
- [Robust Test-time Video-Text Retrieval: Benchmarking and Adapting for Query Shifts](https://mlsanity.com/paper/2604.20851/) (2026)

## Abstract (arXiv)

We review the current schemes of text-image matching models and propose improvements for both training and inference. First, we empirically show limitations of two popular loss (sum and max-margin loss) widely used in training text-image embeddings and propose a trade-off: a kNN-margin loss which 1) utilizes information from hard negatives and 2) is robust to noise as all K-most hardest samples are taken into account, tolerating pseudo negatives and outliers. Second, we advocate the use of Inverted Softmax (textscIs) and Cross-modal Local Scaling (textscCsls) during inference to mitigate the so-called hubness problem in high-dimensional embedding space, enhancing scores of all metrics by a large margin.