Automatic Generation of Contrast Sets from Scene Graphs: Probing the Compositional Consistency of GQA
School of Computer Science and Engineering, The Hebrew University of Jerusalem（耶路撒冷希伯来大学计算机科学与工程学院）；Department of Computer Science, Ben Gurion University（本·古里安大学计算机科学系）
Yonatan Bitton, Gabriel Stanovsky, Roy Schwartz, Michael Elhadad
Recent works have shown that supervised models often exploit data artifacts to achieve good test scores while their performance severely degrades on samples outside their training distribution. Contrast sets (Gardneret al., 2020) quantify this phenomenon by perturbing test samples in a minimal way such that the output label is modified. While most contrast sets were created manually, requiring intensive annotation effort, we present a novel method which leverages rich semantic input representation to automatically generate contrast sets for the visual question answering task. Our method computes the answer of perturbed questions, thus vastly reducing annotation cost and enabling thorough evaluation of models' performance on various semantic aspects (e.g., spatial or relational reasoning). We demonstrate the effectiveness of our approach on the GQA dataset and its semantic scene graph image representation. We find that, despite GQA's compositionality and carefully balanced label distribution, two high-performing models drop 13-17% in accuracy compared to the original test set. Finally, we show that our automatic perturbation can be applied to the training set to mitigate the degradation in performance, opening the door to more robust models.
Multi-Sourced Compositional Generalization in Visual Question Answering
Beijing Key Laboratory of Intelligent Information Technology, School of Computer Science & Technology, Beijing Institute of Technology, China（北京智能信息科技重点实验室，计算机科学与技术学院，北京理工大学，中国）；Guangdong Laboratory of Machine Perception and Intelligent Computing, Shenzhen MSU-BIT University, China（广东机器感知与智能计算实验室，深圳MSU-BIT大学，中国）
With the expressed goal of improving system transparency and visual grounding in the reasoning process in VQA, we present a modular system for the task of compositional VQA based on scene graphs. Our system is called "Adventurer's Treasure Hunt" (or ATH), named after an analogy we draw between our model's search procedure for an answer and an adventurer's search for treasure. We developed ATH with three characteristic features in mind: 1. By design, ATH allows us to explicitly quantify the impact of each of the sub-components on overall VQA performance, as well as their performance on their in…
HyperVis: Continuous Latent Visual Relational Graphs on the Lorentz Hyperboloid for Compositional Reasoning
Data Science and AI, University of Doha for Science and Technology, Qatar（数据科学与人工智能，多哈科学技术大学，卡塔尔）；Pluralis Research, Australia（Pluralis研究，澳大利亚）；Department of Electrical and Computer Engineering, North South University, Bangladesh（电气与计算机工程系，北南大学，孟加拉国）
University of California, Santa Barbara（加州大学圣巴巴拉分校）；Microsoft Dynamics 365 AI Research（微软 Dynamics 365 AI 研究院）
Neural Module Network (NMN) exhibits strong interpretability and compositionality thanks to its handcrafted neural modules with explicit multi-hop reasoning capability. However, most NMNs suffer from two critical drawbacks: 1) scalability: customized module for specific function renders it impractical when scaling up to a larger set of functions in complex tasks; 2) generalizability: rigid pre-defined module inventory makes it difficult to generalize to unseen functions in new tasks/domains. To design a more powerful NMN architecture for practical use, we propose Meta Module Network (MMN) cent…
There are two main lines of research on visual question answering (VQA): compositional model with explicit multi-hop reasoning, and monolithic network with implicit reasoning in the latent feature space. The former excels in interpretability and compositionality but fails on real-world images, while the latter usually achieves better performance due to model flexibility and parameter efficiency. We aim to combine the two to build an interpretable framework for real-world compositional VQA. In our framework, images and questions are disentangled into scene graphs and programs, and a symbolic pr…
GQA~\citep{hudson2019gqa} is a dataset for real-world visual reasoning and compositional question answering. We found that many answers predicted by the best vision-language models on the GQA dataset do not match the ground-truth answer but still are semantically meaningful and correct in the given context. In fact, this is the case with most existing visual question answering (VQA) datasets where they assume only one ground-truth answer for each question. We propose Alternative Answer Sets (AAS) of ground-truth answers to address this limitation, which is created automatically using off-the-s…
Conservatoire National des Arts et Métiers（法国国立工艺学院）；XXII Group（XXII集团）
Visual Question Answering (VQA) is a complex task requiring large datasets and expensive training. Neural Module Networks (NMN) first translate the question to a reasoning path, then follow that path to analyze the image and provide an answer. We propose an NMN method that relies on predefined cross-modal embeddings to ``warm start'' learning on the GQA dataset, then focus on Curriculum Learning (CL) as a way to improve training and make a better use of the data. Several difficulty criteria are employed for defining CL methods. We show that by an appropriate selection of the CL method the cost…
SADL: An Effective In-Context Learning Method for Compositional Visual QA
Posts and Telecommunications Institute of Technology（邮电技术学院）；Deakin University（迪肯大学）；Applied Artificial Intelligence Institute（应用人工智能研究所）；Amazon（亚马逊公司）
We introduce GQA, a new dataset for real-world visual reasoning and compositional question answering, seeking to address key shortcomings of previous VQA datasets. We have developed a strong and robust question engine that leverages scene graph structures to create 22M diverse reasoning questions, all come with functional programs that represent their semantics. We use the programs to gain tight control over the answer distribution and present a new tunable smoothing technique to mitigate question biases. Accompanying the dataset is a suite of new metrics that evaluate essential qualities such…
From Model Patterns to Abstract Semantics in Compositional Zero-Shot Learning
Beijing University of Posts and Telecommunications（北京邮电大学）；Beijing Key Laboratory of Network System and Network Culture（网络系统与网络文化北京市重点实验室）；Key Laboratory of Interactive Technology and Experience System（交互技术与体验系统重点实验室）
Visual question answering (VQA) is a challenging multi-modal task that requires not only the semantic understanding of both images and questions, but also the sound perception of a step-by-step reasoning process that would lead to the correct answer. So far, most successful attempts in VQA have been focused on only one aspect, either the interaction of visual pixel features of images and word features of questions, or the reasoning process of answering the question in an image with simple objects. In this paper, we propose a deep reasoning VQA model with explicit visual structure-aware textual…
Disentanglement-Based Equivariant Learning for Compositional VQA
IEEE Publication Technology Group（IEEE出版技术组）；School of Computing and Artificial Intelligence, Southwest Jiaotong University（计算机与人工智能学院，西南交通大学）；Engineering Research Center of Sustainable Urban Intelligence Transportation, Ministry of Education, China（可持续智慧城市交通工程研究中心，中华人民共和国教育部）；State Key Laboratory of Multimodal Artificial Intelligence Systems (MAIS), Institute of Automation, Chinese Academy of Sciences（多模态人工智能系统（MAIS）国家重点实验室，自动化研究所，中国科学院）；School of Artificial Intelligence, University of Chinese Academy of Sciences（人工智能学院，中国科学院大学）
For vision-and-language reasoning tasks, both fully connectionist, end-to-end methods and hybrid, neuro-symbolic methods have achieved high in-distribution performance. In which out-of-distribution settings does each paradigm excel? We investigate this question on both single-image and multi-image visual question-answering through four types of generalization tests: a novel segment-combine test for multi-image queries, contrast set, compositional generalization, and cross-benchmark transfer. Vision-and-language end-to-end trained systems exhibit sizeable performance drops across all these test…
Visual Question Answering (VQA) is of tremendous interest to the research community with important applications such as aiding visually impaired users and image-based search. In this work, we explore the use of scene graphs for solving the VQA task. We conduct experiments on the GQA dataset which presents a challenging set of questions requiring counting, compositionality and advanced reasoning capability, and provides scene graphs for a large number of images. We adopt image + question architectures for use with scene graphs, evaluate various scene graph generation techniques for unseen image…
We introduce the Neural State Machine, seeking to bridge the gap between the neural and symbolic views of AI and integrate their complementary strengths for the task of visual reasoning. Given an image, we first predict a probabilistic graph that represents its underlying semantics and serves as a structured world model. Then, we perform sequential reasoning over the graph, iteratively traversing its nodes to answer a given question or draw a new inference. In contrast to most neural architectures that are designed to closely interact with the raw sensory data, our model operates instead in an…
Xi’an Jiaotong University（西安交通大学）；Institute of Artificial Intelligence and Robotics（人工智能与机器人研究所）
Encouraging progress has been made towards Visual Question Answering (VQA) in recent years, but it is still challenging to enable VQA models to adaptively generalize to out-of-distribution (OOD) samples. Intuitively, recompositions of existing visual concepts (\ie, attributes and objects) can generate unseen compositions in the training set, which will promote VQA models to generalize to OOD samples. In this paper, we formulate OOD generalization in VQA as a compositional generalization problem and propose a graph generative modeling-based training scheme (X-GGM) to implicitly model the proble…
Hugging Face；Facebook AI Research（脸书人工智能研究院）；University of Waterloo（滑铁卢大学）；University College London（伦敦大学学院）
We present a novel task and dataset for evaluating the ability of vision and language models to conduct visio-linguistic compositional reasoning, which we call Winoground. Given two images and two captions, the goal is to match them correctly - but crucially, both captions contain a completely identical set of words, only in a different order. The dataset was carefully hand-curated by expert annotators and is labeled with a rich set of fine-grained tags to assist in analyzing model performance. We probe a diverse range of state-of-the-art vision and language models and find that, surprisingly,…
Natural language numbers are an example of compositional structures, where larger numbers are composed of operations on smaller numbers. Given that compositional reasoning is a key to natural language understanding, we propose novel multilingual probing tasks tested on DistilBERT, XLM, and BERT to investigate for evidence of compositional reasoning over numerical data in various natural language number systems. By using both grammaticality judgment and value comparison classification tasks in English, Japanese, Danish, and French, we find evidence that the information encoded in these pretrain…
The production of color language is essential for grounded language generation. Color descriptions have many challenging properties: they can be vague, compositionally complex, and denotationally rich. We present an effective approach to generating color descriptions using recurrent neural networks and a Fourier-transformed color representation. Our model outperforms previous work on a conditional language modeling task over a large corpus of naturalistic color descriptions. In addition, probing the model's output reveals that it can accurately produce not only basic color terms but also descr…
Know Yourself and Know Your Neighbour: A Syntactically Informed Self-Supervised Compositional Sentence Representation Learning Framework using a Recursive Hypernetwork
ADAPT Research Centre（ADAPT研究中心）；Trinity College Dublin（都柏林圣三一学院）
Georgia Institute of Technology（佐治亚理工学院）；DeepMind（深度思维）
Recent work has shown how predictive modeling can endow agents with rich knowledge of their surroundings, improving their ability to act in complex environments. We propose question-answering as a general paradigm to decode and understand the representations that such agents develop, applying our method to two recent approaches to predictive modeling -action-conditional CPC (Guo et al., 2018) and SimCore (Gregor et al., 2019). After training agents with these predictive objectives in a visually-rich, 3D environment with an assortment of objects, colors, shapes, and spatial configurations, we p…
Discriminative Probing and Tuning for Text-to-Image Generation
National University of Singapore（新加坡国立大学）；Hong Kong Polytechnic University（香港理工大学）；Nanyang Technological University（南洋理工大学）；Skywork AI（天工AI）；Harbin Institute of Technology (Shenzhen)（哈尔滨工业大学（深圳））
