# What Vision-Language Models `See' when they See Scenes

arXiv: https://arxiv.org/abs/2109.07301 · HTML: https://mlsanity.com/paper/2109.07301/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Placed in VLM benchmarks and visual reasoning, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 34 of its 52 references that are in the map.
- First posted on arXiv 15 Sep 2021.

## Placed in

- Multimodal LLMs / vision-language models / [VLM benchmarks and visual reasoning](https://mlsanity.com/mllm/vlm-benchmarks-and-visual-reasoning/)

## Builds on (most cited first, 30 of 34)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [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)
- [ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks](https://mlsanity.com/paper/1908.02265/) (2019)
- [LXMERT: Learning Cross-Modality Encoder Representations from Transformers](https://mlsanity.com/paper/1908.07490/) (2019)
- [Microsoft COCO Captions: Data Collection and Evaluation Server](https://mlsanity.com/paper/1504.00325/) (2015)
- [Bottom-Up and Top-Down Attention for Image Captioning and Visual Question Answering](https://mlsanity.com/paper/1707.07998/) (2017)
- [VisualBERT: A Simple and Performant Baseline for Vision and Language](https://mlsanity.com/paper/1908.03557/) (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)
- [What Does BERT Look At? An Analysis of BERT's Attention](https://mlsanity.com/paper/1906.04341/) (2019)
- [VL-BERT: Pre-training of Generic Visual-Linguistic Representations](https://mlsanity.com/paper/1908.08530/) (2019)
- [From Recognition to Cognition: Visual Commonsense Reasoning](https://mlsanity.com/paper/1811.10830/) (2018)
- [A Corpus for Reasoning About Natural Language Grounded in Photographs](https://mlsanity.com/paper/1811.00491/) (2018)
- [Unicoder-VL: A Universal Encoder for Vision and Language by Cross-modal Pre-training](https://mlsanity.com/paper/1908.06066/) (2019)
- [SWAG: A Large-Scale Adversarial Dataset for Grounded Commonsense Inference](https://mlsanity.com/paper/1808.05326/) (2018)
- [12-in-1: Multi-Task Vision and Language Representation Learning](https://mlsanity.com/paper/1912.02315/) (2019)
- [Analysis Methods in Neural Language Processing: A Survey](https://mlsanity.com/paper/1812.08951/) (2018)
- [UniVL: A Unified Video and Language Pre-Training Model for Multimodal Understanding and Generation](https://mlsanity.com/paper/2002.06353/) (2020)
- [Experience Grounds Language](https://mlsanity.com/paper/2004.10151/) (2020)
- [Connecting Vision and Language with Localized Narratives](https://mlsanity.com/paper/1912.03098/) (2019)
- [Seeing Out of tHe bOx: End-to-End Pre-training for Vision-Language Representation Learning](https://mlsanity.com/paper/2104.03135/) (2021)
- [Are Convolutional Neural Networks or Transformers more like human vision?](https://mlsanity.com/paper/2105.07197/) (2021)
- [Probing Image-Language Transformers for Verb Understanding](https://mlsanity.com/paper/2106.09141/) (2021)
- [Vision-and-Language or Vision-for-Language? On Cross-Modal Influence in Multimodal Transformers](https://mlsanity.com/paper/2109.04448/) (2021)
- [MiniVLM: A Smaller and Faster Vision-Language Model](https://mlsanity.com/paper/2012.06946/) (2020)
- [GRUEN for Evaluating Linguistic Quality of Generated Text](https://mlsanity.com/paper/2010.02498/) (2020)

## Built on by (most cited first, 8 of 8)

- [Bridging the Digital Divide: Performance Variation across Socio-Economic Factors in Vision-Language Models](https://mlsanity.com/paper/2311.05746/) (2023)
- [Annotations on a Budget: Leveraging Geo-Data Similarity to Balance Model Performance and Annotation Cost](https://mlsanity.com/paper/2403.07687/) (2024)
- [HL Dataset: Visually-grounded Description of Scenes, Actions and Rationales](https://mlsanity.com/paper/2302.12189/) (2023)
- [Why context matters in VQA and Reasoning: Semantic interventions for VLM input modalities](https://mlsanity.com/paper/2410.01690/) (2024)
- [Understanding Cross-modal Interactions in V&L Models that Generate Scene Descriptions](https://mlsanity.com/paper/2211.04971/) (2022)
- [Scalable Performance Analysis for Vision-Language Models](https://mlsanity.com/paper/2305.18786/) (2023)
- [Resilience through Scene Context in Visual Referring Expression Generation](https://mlsanity.com/paper/2404.12289/) (2024)
- [Implementing and Experimenting with Diffusion Models for Text-to-Image Generation](https://mlsanity.com/paper/2209.10948/) (2022)

## Abstract (arXiv)

Images can be described in terms of the objects they contain, or in terms of the types of scene or place that they instantiate. In this paper we address to what extent pretrained Vision and Language models can learn to align descriptions of both types with images. We compare 3 state-of-the-art models, VisualBERT, LXMERT and CLIP. We find that (i) V&L models are susceptible to stylistic biases acquired during pretraining; (ii) only CLIP performs consistently well on both object- and scene-level descriptions. A follow-up ablation study shows that CLIP uses object-level information in the visual modality to align with scene-level textual descriptions.