# TokenFlow: Rethinking Fine-grained Cross-modal Alignment in Vision-Language Retrieval

arXiv: https://arxiv.org/abs/2209.13822 · HTML: https://mlsanity.com/paper/2209.13822/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Builds on 27 of its 37 references that are in the map.
- First posted on arXiv 28 Sep 2022.

## Builds on (most cited first, 27 of 27)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks](https://mlsanity.com/paper/1506.01497/) (2015)
- [Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision](https://mlsanity.com/paper/2102.05918/) (2021)
- [Align before Fuse: Vision and Language Representation Learning with Momentum Distillation](https://mlsanity.com/paper/2107.07651/) (2021)
- [SuperGlue: Learning Feature Matching with Graph Neural Networks](https://mlsanity.com/paper/1911.11763/) (2019)
- [ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision](https://mlsanity.com/paper/2102.03334/) (2021)
- [Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks](https://mlsanity.com/paper/2004.06165/) (2020)
- [Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrieval](https://mlsanity.com/paper/2104.00650/) (2021)
- [Deep Visual-Semantic Alignments for Generating Image Descriptions](https://mlsanity.com/paper/1412.2306/) (2014)
- [Dense-Captioning Events in Videos](https://mlsanity.com/paper/1705.00754/) (2017)
- [CLIP4Clip: An Empirical Study of CLIP for End to End Video Clip Retrieval](https://mlsanity.com/paper/2104.08860/) (2021)
- [FILIP: Fine-grained Interactive Language-Image Pre-Training](https://mlsanity.com/paper/2111.07783/) (2021)
- [Less is More: ClipBERT for Video-and-Language Learning via Sparse Sampling](https://mlsanity.com/paper/2102.06183/) (2021)
- [Stacked Cross Attention for Image-Text Matching](https://mlsanity.com/paper/1803.08024/) (2018)
- [Attention-Based Models for Speech Recognition](https://mlsanity.com/paper/1506.07503/) (2015)
- [Multi-modal Transformer for Video Retrieval](https://mlsanity.com/paper/2007.10639/) (2020)
- [Large-Scale Adversarial Training for Vision-and-Language Representation Learning](https://mlsanity.com/paper/2006.06195/) (2020)
- [An Empirical Study of Training End-to-End Vision-and-Language Transformers](https://mlsanity.com/paper/2111.02387/) (2021)
- [CLIP2Video: Mastering Video-Text Retrieval via Image CLIP](https://mlsanity.com/paper/2106.11097/) (2021)
- [Use What You Have: Video Retrieval Using Representations From Collaborative Experts](https://mlsanity.com/paper/1907.13487/) (2019)
- [A Joint Sequence Fusion Model for Video Question Answering and Retrieval](https://mlsanity.com/paper/1808.02559/) (2018)
- [Improving Video-Text Retrieval by Multi-Stream Corpus Alignment and Dual Softmax Loss](https://mlsanity.com/paper/2109.04290/) (2021)
- [Cross-Modal and Hierarchical Modeling of Video and Text](https://mlsanity.com/paper/1810.07212/) (2018)
- [TACo: Token-aware Cascade Contrastive Learning for Video-Text Alignment](https://mlsanity.com/paper/2108.09980/) (2021)
- [Fine-grained Visual Textual Alignment for Cross-Modal Retrieval using Transformer Encoders](https://mlsanity.com/paper/2008.05231/) (2020)
- [Towards Interpretable Deep Metric Learning with Structural Matching](https://mlsanity.com/paper/2108.05889/) (2021)

## Built on by (most cited first, 6 of 6)

- [Unified Coarse-to-Fine Alignment for Video-Text Retrieval](https://mlsanity.com/paper/2309.10091/) (2023)
- [MLIP: Efficient Multi-Perspective Language-Image Pretraining with Exhaustive Data Utilization](https://mlsanity.com/paper/2406.01460/) (2024)
- [JourneyBench: A Challenging One-Stop Vision-Language Understanding Benchmark of Generated Images](https://mlsanity.com/paper/2409.12953/) (2024)
- [LightCLIP: Learning Multi-Level Interaction for Lightweight Vision-Language Models](https://mlsanity.com/paper/2312.00674/) (2023)
- [Beyond Coarse-Grained Matching in Video-Text Retrieval](https://mlsanity.com/paper/2410.12407/) (2024)
- [EagleNet: Energy-Aware Fine-Grained Relationship Learning Network for Text-Video Retrieval](https://mlsanity.com/paper/2603.25267/) (2026)

## Abstract (arXiv)

Most existing methods in vision-language retrieval match two modalities by either comparing their global feature vectors which misses sufficient information and lacks interpretability, detecting objects in images or videos and aligning the text with fine-grained features which relies on complicated model designs, or modeling fine-grained interaction via cross-attention upon visual and textual tokens which suffers from inferior efficiency. To address these limitations, some recent works simply aggregate the token-wise similarities to achieve fine-grained alignment, but they lack intuitive explanations as well as neglect the relationships between token-level features and global representations with high-level semantics. In this work, we rethink fine-grained cross-modal alignment and devise a new model-agnostic formulation for it. We additionally demystify the recent popular works and subsume them into our scheme. Furthermore, inspired by optimal transport theory, we introduce TokenFlow, an instantiation of the proposed scheme. By modifying only the similarity function, the performance of our method is comparable to the SoTA algorithms with heavy model designs on major video-text retrieval benchmarks. The visualization further indicates that TokenFlow successfully leverages the fine-grained information and achieves better interpretability.