# RETTA: Retrieval-Enhanced Test-Time Adaptation for Zero-Shot Video Captioning

arXiv: https://arxiv.org/abs/2405.07046 · HTML: https://mlsanity.com/paper/2405.07046/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 30 of its 50 references that are in the map.
- First posted on arXiv 11 May 2024.

## Builds on (most cited first, 30 of 30)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Dense Passage Retrieval for Open-Domain Question Answering](https://mlsanity.com/paper/2004.04906/) (2020)
- [REALM: Retrieval-Augmented Language Model Pre-Training](https://mlsanity.com/paper/2002.08909/) (2020)
- [Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video Understanding](https://mlsanity.com/paper/2306.02858/) (2023)
- [Microsoft COCO Captions: Data Collection and Evaluation Server](https://mlsanity.com/paper/1504.00325/) (2015)
- [Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivity](https://mlsanity.com/paper/2104.08786/) (2021)
- [In-Context Retrieval-Augmented Language Models](https://mlsanity.com/paper/2302.00083/) (2023)
- [Latent Retrieval for Weakly Supervised Open Domain Question Answering](https://mlsanity.com/paper/1906.00300/) (2019)
- [CLIP4Clip: An Empirical Study of CLIP for End to End Video Clip Retrieval](https://mlsanity.com/paper/2104.08860/) (2021)
- [When and why vision-language models behave like bags-of-words, and what to do about it?](https://mlsanity.com/paper/2210.01936/) (2022)
- [VATEX: A Large-Scale, High-Quality Multilingual Dataset for Video-and-Language Research](https://mlsanity.com/paper/1904.03493/) (2019)
- [Test-Time Prompt Tuning for Zero-Shot Generalization in Vision-Language Models](https://mlsanity.com/paper/2209.07511/) (2022)
- [Do Prompt-Based Models Really Understand the Meaning of their Prompts?](https://mlsanity.com/paper/2109.01247/) (2021)
- [Expanding Language-Image Pretrained Models for General Video Recognition](https://mlsanity.com/paper/2208.02816/) (2022)
- [RECOMP: Improving Retrieval-Augmented LMs with Compression and Selective Augmentation](https://mlsanity.com/paper/2310.04408/) (2023)
- [ZeroCap: Zero-Shot Image-to-Text Generation for Visual-Semantic Arithmetic](https://mlsanity.com/paper/2111.14447/) (2021)
- [Language Models Can See: Plugging Visual Controls in Text Generation](https://mlsanity.com/paper/2205.02655/) (2022)
- [WenLan: Bridging Vision and Language by Large-Scale Multi-Modal Pre-Training](https://mlsanity.com/paper/2103.06561/) (2021)
- [DeCap: Decoding CLIP Latents for Zero-Shot Captioning via Text-Only Training](https://mlsanity.com/paper/2303.03032/) (2023)
- [Open-book Video Captioning with Retrieve-Copy-Generate Network](https://mlsanity.com/paper/2103.05284/) (2021)
- [Cross-Modal Adapter for Vision-Language Retrieval](https://mlsanity.com/paper/2211.09623/) (2022)
- [MeaCap: Memory-Augmented Zero-shot Image Captioning](https://mlsanity.com/paper/2403.03715/) (2024)
- [Hierarchical Modular Network for Video Captioning](https://mlsanity.com/paper/2111.12476/) (2021)
- [Zero-Shot Video Captioning with Evolving Pseudo-Tokens](https://mlsanity.com/paper/2207.11100/) (2022)
- [Comprehensive Information Integration Modeling Framework for Video Titling](https://mlsanity.com/paper/2006.13608/) (2020)
- [Delving Deeper into the Decoder for Video Captioning](https://mlsanity.com/paper/2001.05614/) (2020)
- [MultiCapCLIP: Auto-Encoding Prompts for Zero-Shot Multilingual Visual Captioning](https://mlsanity.com/paper/2308.13218/) (2023)
- [VPA: Fully Test-Time Visual Prompt Adaptation](https://mlsanity.com/paper/2309.15251/) (2023)
- [Few-shot Transfer Learning for Knowledge Base Question Answering: Fusing Supervised Models with In-Context Learning](https://mlsanity.com/paper/2311.08894/) (2023)
- [Zero-Shot Dense Video Captioning by Jointly Optimizing Text and Moment](https://mlsanity.com/paper/2307.02682/) (2023)

## Built on by (most cited first, 5 of 5)

- [AuroraCap: Efficient, Performant Video Detailed Captioning and a New Benchmark](https://mlsanity.com/paper/2410.03051/) (2024)
- [Language-Guided Graph Representation Learning for Video Summarization](https://mlsanity.com/paper/2511.10953/) (2025)
- [Bridging Vision and Language: Modeling Causality and Temporality in Video Narratives](https://mlsanity.com/paper/2412.10720/) (2024)
- [Exploring the Temporal Consistency for Point-Level Weakly-Supervised Temporal Action Localization](https://mlsanity.com/paper/2602.05718/) (2026)
- [Watching Synthetic Videos: Aligning Cross-modal Representations with Visual Synthesis for Zero-shot Video Captioning](https://mlsanity.com/paper/2608.11013/) (2026)

## Abstract (arXiv)

Despite the significant progress of fully-supervised video captioning, zero-shot methods remain much less explored. In this paper, we propose a novel zero-shot video captioning framework named Retrieval-Enhanced Test-Time Adaptation (RETTA), which takes advantage of existing pretrained large-scale vision and language models to directly generate captions with test-time adaptation. Specifically, we bridge video and text using four key models: a general video-text retrieval model XCLIP, a general image-text matching model CLIP, a text alignment model AnglE, and a text generation model GPT-2, due to their source-code availability. The main challenge is how to enable the text generation model to be sufficiently aware of the content in a given video so as to generate corresponding captions. To address this problem, we propose using learnable tokens as a communication medium among these four frozen models GPT-2, XCLIP, CLIP, and AnglE. Different from the conventional way that trains these tokens with training data, we propose to learn these tokens with soft targets of the inference data under several carefully crafted loss functions, which enable the tokens to absorb video information catered for GPT-2. This procedure can be efficiently done in just a few iterations (we use 16 iterations in the experiments) and does not require ground truth data. Extensive experimental results on three widely used datasets, MSR-VTT, MSVD, and VATEX, show absolute 5.1%-32.4% improvements in terms of the main metric CIDEr compared to several state-of-the-art zero-shot video captioning methods.