Caption Anything: Interactive Image Description with Diverse Multimodal Controls

Controllable image captioning is an emerging multimodal topic that aims to describe the image with natural language following human purpose, $\textit{e.g.}$, looking at the specified regions or telling in a particular text style. State-of-the-art methods are trained on annotated pairs of input controls and output captions. However, the scarcity of such well-annotated multimodal data largely limits their usability and scalability for interactive AI systems. Leveraging unimodal instruction-following foundation models is a promising alternative that benefits from broader sources of data. In this paper, we present Caption AnyThing (CAT), a foundation model augmented image captioning framework supporting a wide range of multimodel controls: 1) visual controls, including points, boxes, and trajectories; 2) language controls, such as sentiment, length, language, and factuality. Powered by Segment Anything Model (SAM) and ChatGPT, we unify the visual and language prompts into a modularized framework, enabling the flexible combination between different controls. Extensive case studies demonstrate the user intention alignment capabilities of our framework, shedding light on effective user interaction modeling in vision-language applications. Our code is publicly available at https://github.com/ttengwang/Caption-Anything.

[1]  Ying Shan,et al.  Accelerating Vision-Language Pretraining with Free Language Modeling , 2023, 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[2]  Henrique Pondé de Oliveira Pinto,et al.  GPT-4 Technical Report , 2023, 2303.08774.

[3]  Chenfei Wu,et al.  Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation Models , 2023, ArXiv.

[4]  Zhengjue Wang,et al.  ConZIC: Controllable Zero-shot Image Captioning by Sampling-Based Polishing , 2023, 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[5]  Naman Goyal,et al.  LLaMA: Open and Efficient Foundation Language Models , 2023, ArXiv.

[6]  S. Savarese,et al.  BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models , 2023, ArXiv.

[7]  Jiahao Xie,et al.  Controllable Image Captioning via Prompting , 2022, AAAI.

[8]  Zhe Gan,et al.  GRiT: A Generative Region-to-text Transformer for Object Understanding , 2022, ArXiv.

[9]  Guillem Cucurull,et al.  Galactica: A Large Language Model for Science , 2022, ArXiv.

[10]  Alexander M. Rush,et al.  BLOOM: A 176B-Parameter Open-Access Multilingual Language Model , 2022, ArXiv.

[11]  Dragomir R. Radev,et al.  Crosslingual Generalization through Multitask Finetuning , 2022, ArXiv.

[12]  Andrew M. Dai,et al.  Scaling Instruction-Finetuned Language Models , 2022, ArXiv.

[13]  Q. Tian,et al.  DeeCap: Dynamic Early Exiting for Efficient Image Captioning , 2022, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[14]  Ting Yao,et al.  Comprehending and Ordering Semantics for Image Captioning , 2022, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[15]  Zhe Gan,et al.  GIT: A Generative Image-to-text Transformer for Vision and Language , 2022, Trans. Mach. Learn. Res..

[16]  S. Gu,et al.  Large Language Models are Zero-Shot Reasoners , 2022, NeurIPS.

[17]  Xi Victoria Lin,et al.  OPT: Open Pre-trained Transformer Language Models , 2022, ArXiv.

[18]  Kurt Debattista,et al.  Region-Object Relation-Aware Dense Captioning via Transformer. , 2022, IEEE transactions on neural networks and learning systems.

[19]  Ryan J. Lowe,et al.  Training language models to follow instructions with human feedback , 2022, NeurIPS.

[20]  S. Hoi,et al.  BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation , 2022, ICML.

[21]  Dale Schuurmans,et al.  Chain of Thought Prompting Elicits Reasoning in Large Language Models , 2022, NeurIPS.

[22]  Xiaowei Hu,et al.  Injecting Semantic Concepts into End-to-End Image Captioning , 2021, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[23]  Xiaowei Hu,et al.  Scaling Up Vision-Language Pretraining for Image Captioning , 2021, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[24]  Wei Liu,et al.  Human-like Controllable Image Captioning with Verb-specific Semantic Roles , 2021, 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[25]  Nan Duan,et al.  Control Image Captioning Spatially and Temporally , 2021, ACL.

[26]  Ning Ding,et al.  Length-Controllable Image Captioning , 2020, ECCV.

[27]  Zhao Zhang,et al.  Interactive Image Segmentation With First Click Attention , 2020, 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[28]  Mark Chen,et al.  Language Models are Few-Shot Learners , 2020, NeurIPS.

[29]  Wentian Zhao,et al.  MemCap: Memorizing Style Knowledge for Image Captioning , 2020, AAAI.

[30]  Ilia Petrov,et al.  F-BRS: Rethinking Backpropagating Refinement for Interactive Segmentation , 2020, 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[31]  Jordi Pont-Tuset,et al.  Connecting Vision and Language with Localized Narratives , 2019, ECCV.

[32]  Jie Chen,et al.  Attention on Attention for Image Captioning , 2019, 2019 IEEE/CVF International Conference on Computer Vision (ICCV).

[33]  Jungong Han,et al.  Learning Object Context for Dense Captioning , 2019, AAAI.

[34]  Nenghai Yu,et al.  Context and Attribute Grounded Dense Captioning , 2019, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[35]  Rita Cucchiara,et al.  Show, Control and Tell: A Framework for Generating Controllable and Grounded Captions , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[36]  Sébastien Ourselin,et al.  DeepIGeoS: A Deep Interactive Geodesic Framework for Medical Image Segmentation , 2017, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[37]  Zhuwen Li,et al.  Interactive Image Segmentation with Latent Diversity , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[38]  Bastian Leibe,et al.  Iteratively Trained Interactive Segmentation , 2018, BMVC.

[39]  Lei Zhang,et al.  Bottom-Up and Top-Down Attention for Image Captioning and Visual Question Answering , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[40]  Sim Heng Ong,et al.  Regional Interactive Image Segmentation Networks , 2017, 2017 IEEE International Conference on Computer Vision (ICCV).

[41]  Zhe Gan,et al.  StyleNet: Generating Attractive Visual Captions with Styles , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[42]  Li-Jia Li,et al.  Dense Captioning with Joint Inference and Visual Context , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[43]  Ning Xu,et al.  Deep Interactive Object Selection , 2016, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[44]  Li Fei-Fei,et al.  DenseCap: Fully Convolutional Localization Networks for Dense Captioning , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[45]  Lexing Xie,et al.  SentiCap: Generating Image Descriptions with Sentiments , 2015, AAAI.

[46]  Yoshua Bengio,et al.  Show, Attend and Tell: Neural Image Caption Generation with Visual Attention , 2015, ICML.

[47]  Samy Bengio,et al.  Show and tell: A neural image caption generator , 2014, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).