# Context-Aware Visual Policy Network for Sequence-Level Image Captioning

arXiv: https://arxiv.org/abs/1808.05864 · HTML: https://mlsanity.com/paper/1808.05864/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 16 Aug 2018.

## Built on by (most cited first, 12 of 12)

- [Causal Attention for Vision-Language Tasks](https://mlsanity.com/paper/2103.03493/) (2021)
- [Semantic-Conditional Diffusion Networks for Image Captioning](https://mlsanity.com/paper/2212.03099/) (2022)
- [EMScore: Evaluating Video Captioning via Coarse-Grained and Fine-Grained Embedding Matching](https://mlsanity.com/paper/2111.08919/) (2021)
- [DiMBERT: Learning Vision-Language Grounded Representations with Disentangled Multimodal-Attention](https://mlsanity.com/paper/2210.16431/) (2022)
- [Label-Attention Transformer with Geometrically Coherent Objects for Image Captioning](https://mlsanity.com/paper/2109.07799/) (2021)
- [Edit As You Wish: Video Caption Editing with Multi-grained User Control](https://mlsanity.com/paper/2305.08389/) (2023)
- [Learning to Collocate Visual-Linguistic Neural Modules for Image Captioning](https://mlsanity.com/paper/2210.01338/) (2022)
- [Memory Enhanced Embedding Learning for Cross-Modal Video-Text Retrieval](https://mlsanity.com/paper/2103.15686/) (2021)
- [Goal-driven text descriptions for images](https://mlsanity.com/paper/2108.12575/) (2021)
- [How to Describe Images in a More Funny Way? Towards a Modular Approach to Cross-Modal Sarcasm Generation](https://mlsanity.com/paper/2211.10992/) (2022)
- [OSIC: A New One-Stage Image Captioner Coined](https://mlsanity.com/paper/2211.02321/) (2022)
- [A Fine-Grained Image Description Generation Method Based on Joint Objectives](https://mlsanity.com/paper/2311.12799/) (2023)

## Abstract (arXiv)

Many vision-language tasks can be reduced to the problem of sequence prediction for natural language output. In particular, recent advances in image captioning use deep reinforcement learning (RL) to alleviate the "exposure bias" during training: ground-truth subsequence is exposed in every step prediction, which introduces bias in test when only predicted subsequence is seen. However, existing RL-based image captioning methods only focus on the language policy while not the visual policy (e.g., visual attention), and thus fail to capture the visual context that are crucial for compositional reasoning such as visual relationships (e.g., "man riding horse") and comparisons (e.g., "smaller cat"). To fill the gap, we propose a Context-Aware Visual Policy network (CAVP) for sequence-level image captioning. At every time step, CAVP explicitly accounts for the previous visual attentions as the context, and then decides whether the context is helpful for the current word generation given the current visual attention. Compared against traditional visual attention that only fixes a single image region at every step, CAVP can attend to complex visual compositions over time. The whole image captioning model --- CAVP and its subsequent language policy network --- can be efficiently optimized end-to-end by using an actor-critic policy gradient method with respect to any caption evaluation metric. We demonstrate the effectiveness of CAVP by state-of-the-art performances on MS-COCO offline split and online server, using various metrics and sensible visualizations of qualitative visual context. The code is available at https://github.com/daqingliu/CAVP