The field of Abstract Visual Reasoning (AVR) encompasses a wide range of problems, many of which are inspired by human IQ tests. The variety of AVR tasks has resulted in state-of-the-art AVR methods being task-specific approaches. Furthermore, contemporary methods consider each AVR problem instance not as a whole, but in the form of a set of individual panels with particular locations and roles (context vs. answer panels) pre-assigned according to the task-specific arrangements. While these highly specialized approaches have recently led to significant progress in solving particular AVR tasks, considering each task in isolation hinders the development of universal learning systems in this domain. In this paper, we introduce a unified view of AVR tasks, where each problem instance is rendered as a single image, with no a priori assumptions about the number of panels, their location, or role. The main advantage of the proposed unified view is the ability to develop universal learning models applicable to various AVR tasks. What is more, the proposed approach inherently facilitates transfer learning in the AVR domain, as various types of problems share a common representation. The experiments conducted on four AVR datasets with Raven's Progressive Matrices and Visual Analogy Problems, and one real-world visual analogy dataset show that the proposed unified representation of AVR tasks poses a challenge to state-of-the-art Deep Learning (DL) AVR models and, more broadly, contemporary DL image recognition methods. In order to address this challenge, we introduce the Unified Model for Abstract Visual Reasoning (UMAVR) capable of dealing with various types of AVR problems in a unified manner. UMAVR outperforms existing AVR methods in selected single-task learning experiments, and demonstrates effective knowledge reuse in transfer learning and curriculum learning setups.
We study generalization and knowledge reuse capabilities of deep neural networks in the domain of abstract visual reasoning (AVR), employing Raven's Progressive Matrices (RPMs), a recognized benchmark task for assessing AVR abilities. Two knowledge transfer scenarios referring to the I-RAVEN dataset are investigated. Firstly, inspired by generalization assessment capabilities of the PGM dataset and popularity of I-RAVEN, we introduce Attributeless-I-RAVEN (A-I-RAVEN), a benchmark with 10 generalization regimes that allow to systematically test generalization of abstract rules applied to held-out attributes at various levels of complexity (primary and extended regimes). In contrast to PGM, A-I-RAVEN features compositionality, a variety of figure configurations, and does not require substantial computational resources. Secondly, we construct I-RAVEN-Mesh, a dataset that enriches RPMs with a novel component structure comprising line-based patterns, facilitating assessment of progressive knowledge acquisition in transfer learning setting. We evaluate 13 strong models from the AVR literature on the introduced datasets, revealing their specific shortcomings in generalization and knowledge transfer.
The challenge in object-based visual reasoning lies in generating concept representations that are both descriptive and distinct. Achieving this in an unsupervised manner requires human users to understand the model's learned concepts and, if necessary, revise incorrect ones. To address this challenge, we introduce the Neural Concept Binder (NCB), a novel framework for deriving both discrete and continuous concept representations, which we refer to as "concept-slot encodings". NCB employs two types of binding: "soft binding", which leverages the recent SysBinder mechanism to obtain object-factor encodings, and subsequent "hard binding", achieved through hierarchical clustering and retrieval-based inference. This enables obtaining expressive, discrete representations from unlabeled images. Moreover, the structured nature of NCB's concept representations allows for intuitive inspection and the straightforward integration of external knowledge, such as human input or insights from other AI models like GPT-4. Additionally, we demonstrate that incorporating the hard binding mechanism preserves model performance while enabling seamless integration into both neural and symbolic modules for complex reasoning tasks. We validate the effectiveness of NCB through evaluations on our newly introduced CLEVR-Sudoku dataset.
Comparing two images in terms of Commonalities and Differences (CaD) is a fundamental human capability that forms the basis of advanced visual reasoning and interpretation. It is essential for the generation of detailed and contextually relevant descriptions, performing comparative analysis, novelty detection, and making informed decisions based on visual data. However, surprisingly, little attention has been given to these fundamental concepts in the best current mimic of human visual intelligence - Large Multimodal Models (LMMs). We develop and contribute a new two-phase approach CaD-VI for collecting synthetic visual instructions, together with an instruction-following dataset CaD-Inst containing 349K image pairs with CaD instructions collected using CaD-VI. Our approach significantly improves the CaD spotting capabilities in LMMs, advancing the SOTA on a diverse set of related tasks by up to 17.5%. It is also complementary to existing difference-only instruction datasets, allowing automatic targeted refinement of those resources increasing their effectiveness for CaD tuning by up to 10%. Additionally, we propose an evaluation benchmark with 7.5K open-ended QAs to assess the CaD understanding abilities of LMMs.
Humans draw to facilitate reasoning: we draw auxiliary lines when solving geometry problems; we mark and circle when reasoning on maps; we use sketches to amplify our ideas and relieve our limited-capacity working memory. However, such actions are missing in current multimodal language models (LMs). Current chain-of-thought and tool-use paradigms only use text as intermediate reasoning steps. In this work, we introduce Sketchpad, a framework that gives multimodal LMs a visual sketchpad and tools to draw on the sketchpad. The LM conducts planning and reasoning according to the visual artifacts it has drawn. Different from prior work, which uses text-to-image models to enable LMs to draw, Sketchpad enables LMs to draw with lines, boxes, marks, etc., which is closer to human sketching and better facilitates reasoning. Sketchpad can also use specialist vision models during the sketching process (e.g., draw bounding boxes with object detection models, draw masks with segmentation models), to further enhance visual perception and reasoning. We experiment with a wide range of math tasks (including geometry, functions, graphs, and chess) and complex visual reasoning tasks. Sketchpad substantially improves performance on all tasks over strong base models with no sketching, yielding an average gain of 12.7% on math tasks, and 8.6% on vision tasks. GPT-4o with Sketchpad sets a new state of the art on all tasks, including V*Bench (80.3%), BLINK spatial reasoning (83.9%), and visual correspondence (80.8%). All codes and data are in https://visualsketchpad.github.io/.
Large Vision-Language Models (LVLMs) have demonstrated outstanding performance in various general multimodal applications such as image recognition and visual reasoning, and have also shown promising potential in specialized domains. However, the application potential of LVLMs in the insurance domain-characterized by rich application scenarios and abundant multimodal data-has not been effectively explored. There is no systematic review of multimodal tasks in the insurance domain, nor a benchmark specifically designed to evaluate the capabilities of LVLMs in insurance. This gap hinders the development of LVLMs within the insurance domain. In this paper, we systematically review and distill multimodal tasks for four representative types of insurance: auto insurance, property insurance, health insurance, and agricultural insurance. We propose INS-MMBench, the first comprehensive LVLMs benchmark tailored for the insurance domain. INS-MMBench comprises a total of 2.2K thoroughly designed multiple-choice questions, covering 12 meta-tasks and 22 fundamental tasks. Furthermore, we evaluate multiple representative LVLMs, including closed-source models such as GPT-4o and open-source models like BLIP-2. This evaluation not only validates the effectiveness of our benchmark but also provides an in-depth performance analysis of current LVLMs on various multimodal tasks in the insurance domain. We hope that INS-MMBench will facilitate the further application of LVLMs in the insurance domain and inspire interdisciplinary development. Our dataset and evaluation code are available at https://github.com/FDU-INS/INS-MMBench.
Multimodal Large Language Models (MLLMs) have demonstrated proficiency in processing di-verse modalities, including text, images, and audio. These models leverage extensive pre-existing knowledge, enabling them to address complex problems with minimal to no specific training examples, as evidenced in few-shot and zero-shot in-context learning scenarios. This paper investigates the use of MLLMs' visual capabilities to 'eyeball' solutions for the Traveling Salesman Problem (TSP) by analyzing images of point distributions on a two-dimensional plane. Our experiments aimed to validate the hypothesis that MLLMs can effectively 'eyeball' viable TSP routes. The results from zero-shot, few-shot, self-ensemble, and self-refine zero-shot evaluations show promising outcomes. We anticipate that these findings will inspire further exploration into MLLMs' visual reasoning abilities to tackle other combinatorial problems.
Current video-language models (VLMs) rely extensively on instance-level alignment between video and language modalities, which presents two major limitations: (1) visual reasoning disobeys the natural perception that humans do in first-person perspective, leading to a lack of reasoning interpretation; and (2) learning is limited in capturing inherent fine-grained relationships between two modalities. In this paper, we take an inspiration from human perception and explore a compositional approach for egocentric video representation. We introduce HENASY (Hierarchical ENtities ASsemblY), which includes a spatiotemporal token grouping mechanism to explicitly assemble dynamically evolving scene entities through time and model their relationship for video representation. By leveraging compositional structure understanding, HENASY possesses strong interpretability via visual grounding with free-form text queries. We further explore a suite of multi-grained contrastive losses to facilitate entity-centric understandings. This comprises three alignment types: video-narration, noun-entity, verb-entities alignments. Our method demonstrates strong interpretability in both quantitative and qualitative experiments; while maintaining competitive performances on five downstream tasks via zero-shot transfer or as video/text representation, including video/text retrieval, action recognition, multi-choice query, natural language query, and moments query.
Human visual reasoning is characterized by an ability to identify abstract patterns from only a small number of examples, and to systematically generalize those patterns to novel inputs. This capacity depends in large part on our ability to represent complex visual inputs in terms of both objects and relations. Recent work in computer vision has introduced models with the capacity to extract object-centric representations, leading to the ability to process multi-object visual inputs, but falling short of the systematic generalization displayed by human reasoning. Other recent models have employed inductive biases for relational abstraction to achieve systematic generalization of learned abstract rules, but have generally assumed the presence of object-focused inputs. Here, we combine these two approaches, introducing Object-Centric Relational Abstraction (OCRA), a model that extracts explicit representations of both objects and abstract relations, and achieves strong systematic generalization in tasks (including a novel dataset, CLEVR-ART, with greater visual complexity) involving complex visual displays.
Iteratively improving and repairing source code with large language models (LLMs), known as refinement, has emerged as a popular way of generating programs that would be too complex to construct in one shot. Given a bank of test cases, together with a candidate program, an LLM can improve that program by being prompted with failed test cases. But it remains an open question how to best iteratively refine code, with prior work employing simple greedy or breadth-first strategies. We show here that refinement exposes an explore-exploit tradeoff: exploit by refining the program that passes the most test cases, or explore by refining a lesser considered program. We frame this as an arm-acquiring bandit problem, which we solve with Thompson Sampling. The resulting LLM-based program synthesis algorithm is broadly applicable: Across loop invariant synthesis, visual reasoning puzzles, and competition programming problems, we find that our new method can solve more problems using fewer language model calls.
Recent advancements in Multi-modal Large Language Models (MLLMs) have significantly improved their performance in tasks combining vision and language. However, challenges persist in detailed multi-modal understanding, comprehension of complex tasks, and reasoning over multi-modal information. This paper introduces MMCTAgent, a novel multi-modal critical thinking agent framework designed to address the inherent limitations of current MLLMs in complex visual reasoning tasks. Inspired by human cognitive processes and critical thinking, MMCTAgent iteratively analyzes multi-modal information, decomposes queries, plans strategies, and dynamically evolves its reasoning. Additionally, MMCTAgent incorporates critical thinking elements such as verification of final answers and self-reflection through a novel approach that defines a vision-based critic and identifies task-specific evaluation criteria, thereby enhancing its decision-making abilities. Through rigorous evaluations across various image and video understanding benchmarks, we demonstrate that MMCTAgent (with and without the critic) outperforms both foundational MLLMs and other tool-augmented pipelines.
Visual Commonsense Reasoning (VCR) calls for explanatory reasoning behind question answering over visual scenes. To achieve this goal, a model is required to provide an acceptable rationale as the reason for the predicted answers. Progress on the benchmark dataset stems largely from the recent advancement of Vision-Language Transformers (VL Transformers). These models are first pre-trained on some generic large-scale vision-text datasets, and then the learned representations are transferred to the downstream VCR task. Despite their attractive performance, this paper posits that the VL Transformers do not exhibit visual commonsense, which is the key to VCR. In particular, our empirical results pinpoint several shortcomings of existing VL Transformers: small gains from pre-training, unexpected language bias, limited model architecture for the two inseparable sub-tasks, and neglect of the important object-tag correlation. With these findings, we tentatively suggest some future directions from the aspect of dataset, evaluation metric, and training tricks. We believe this work could make researchers revisit the intuition and goals of VCR, and thus help tackle the remaining challenges in visual reasoning.
Large Vision-Language Models (LVLMs) often produce responses that misalign with factual information, a phenomenon known as hallucinations. While hallucinations are well-studied, the exact causes behind them remain underexplored. In this paper, we first investigate the root causes of hallucinations in LVLMs. Our findings reveal that existing mitigation techniques primarily reduce hallucinations for visual recognition prompts-those that require simple descriptions of visual elements-but fail for cognitive prompts that demand deliberate reasoning. We identify the core issue as a lack of true visual perception in LVLMs: although they can accurately recognize visual elements, they struggle to fully interpret these elements in the context of the input prompt and effectively link this recognition to their internal knowledge, which is critical for reasoning. To address this gap, we introduce Visual Description Grounded Decoding (VDGD), a simple, robust, and training-free method designed to enhance visual perception and improve reasoning capabilities in LVLMs. VDGD works by first generating a detailed description of the image and appending it as a prefix to the instruction. During response generation, tokens are sampled based on their KL divergence to the description, favoring candidates with lower divergence. Experimental results on multiple visual reasoning benchmarks and LVLMs demonstrate that VDGD consistently outperforms existing baselines 2% - 33%. Finally, we introduce VaLLu, a benchmark designed for comprehensive evaluation of the cognitive capabilities of LVLMs.
Recent advancements in Chain-of-Thought (CoT) and related rationale-based works have significantly improved the performance of Large Language Models (LLMs) in complex reasoning tasks. With the evolution of Multimodal Large Language Models (MLLMs), enhancing their capability to tackle complex multimodal reasoning problems is a crucial frontier. However, incorporating multimodal rationales in CoT has yet to be thoroughly investigated. We propose the Image-of-Thought (IoT) prompting method, which helps MLLMs to extract visual rationales step-by-step. Specifically, IoT prompting can automatically design critical visual information extraction operations based on the input images and questions. Each step of visual information refinement identifies specific visual rationales that support answers to complex visual reasoning questions. Beyond the textual CoT, IoT simultaneously utilizes visual and textual rationales to help MLLMs understand complex multimodal information. IoT prompting has improved zero-shot visual reasoning performance across various visual understanding tasks in different MLLMs. Moreover, the step-by-step visual feature explanations generated by IoT prompting elucidate the visual reasoning process, aiding in analyzing the cognitive processes of large multimodal models
Visual In-Context Learning (ICL) has emerged as a promising research area due to its capability to accomplish various tasks with limited example pairs through analogical reasoning. However, training-based visual ICL has limitations in its ability to generalize to unseen tasks and requires the collection of a diverse task dataset. On the other hand, existing methods in the inference-based visual ICL category solely rely on textual prompts, which fail to capture fine-grained contextual information from given examples and can be time-consuming when converting from images to text prompts. To address these challenges, we propose Analogist, a novel inference-based visual ICL approach that exploits both visual and textual prompting techniques using a text-to-image diffusion model pretrained for image inpainting. For visual prompting, we propose a self-attention cloning (SAC) method to guide the fine-grained structural-level analogy between image examples. For textual prompting, we leverage GPT-4V's visual reasoning capability to efficiently generate text prompts and introduce a cross-attention masking (CAM) operation to enhance the accuracy of semantic-level analogy guided by text prompts. Our method is out-of-the-box and does not require fine-tuning or optimization. It is also generic and flexible, enabling a wide range of visual tasks to be performed in an in-context manner. Extensive experiments demonstrate the superiority of our method over existing approaches, both qualitatively and quantitatively.
While vision-language pretrained models (VLMs) excel in various multimodal understanding tasks, their potential in fine-grained audio-visual reasoning, particularly for audio-visual question answering (AVQA), remains largely unexplored. AVQA presents specific challenges for VLMs due to the requirement of visual understanding at the region level and seamless integration with audio modality. Previous VLM-based AVQA methods merely used CLIP as a feature encoder but underutilized its knowledge, and mistreated audio and video as separate entities in a dual-stream framework as most AVQA methods. This paper proposes a new CLIP-powered target-aware single-stream (TASS) network for AVQA using the image-text matching knowledge of the pretrained model through the audio-visual matching characteristic of nature. It consists of two key components: the target-aware spatial grounding module (TSG+) and the single-stream joint temporal grounding module (JTG). Specifically, we propose a TSG+ module to transfer the image-text matching knowledge from CLIP models to our region-text matching process without corresponding ground-truth labels. Moreover, unlike previous separate dual-stream networks that still required an additional audio-visual fusion module, JTG unifies audio-visual fusion and question-aware temporal grounding in a simplified single-stream architecture. It treats audio and video as a cohesive entity and further extends the pretrained image-text knowledge to audio-text matching by preserving their temporal correlation with our proposed cross-modal synchrony (CMS) loss. Extensive experiments conducted on the MUSIC-AVQA benchmark verified the effectiveness of our proposed method over existing state-of-the-art methods.
Learning compositional representation is a key aspect of object-centric learning as it enables flexible systematic generalization and supports complex visual reasoning. However, most of the existing approaches rely on auto-encoding objective, while the compositionality is implicitly imposed by the architectural or algorithmic bias in the encoder. This misalignment between auto-encoding objective and learning compositionality often results in failure of capturing meaningful object representations. In this study, we propose a novel objective that explicitly encourages compositionality of the representations. Built upon the existing object-centric learning framework (e.g., slot attention), our method incorporates additional constraints that an arbitrary mixture of object representations from two images should be valid by maximizing the likelihood of the composite data. We demonstrate that incorporating our objective to the existing framework consistently improves the objective-centric learning and enhances the robustness to the architectural choices.
3D visual grounding is a challenging task that often requires direct and dense supervision, notably the semantic label for each object in the scene. In this paper, we instead study the naturally supervised setting that learns from only 3D scene and QA pairs, where prior works underperform. We propose the Language-Regularized Concept Learner (LARC), which uses constraints from language as regularization to significantly improve the accuracy of neuro-symbolic concept learners in the naturally supervised setting. Our approach is based on two core insights: the first is that language constraints (e.g., a word's relation to another) can serve as effective regularization for structured representations in neuro-symbolic models; the second is that we can query large language models to distill such constraints from language properties. We show that LARC improves performance of prior works in naturally supervised 3D visual grounding, and demonstrates a wide range of 3D visual reasoning capabilities-from zero-shot composition, to data efficiency and transferability. Our method represents a promising step towards regularizing structured visual reasoning frameworks with language-based priors, for learning in settings without dense supervision.
Graphics design is important for various applications, including movie production and game design. To create a high-quality scene, designers usually need to spend hours in software like Blender, in which they might need to interleave and repeat operations, such as connecting material nodes, hundreds of times. Moreover, slightly different design goals may require completely different sequences, making automation difficult. In this paper, we propose a system that leverages Vision-Language Models (VLMs), like GPT-4V, to intelligently search the design action space to arrive at an answer that can satisfy a user's intent. Specifically, we design a vision-based edit generator and state evaluator to work together to find the correct sequence of actions to achieve the goal. Inspired by the role of visual imagination in the human design process, we supplement the visual reasoning capabilities of VLMs with "imagined" reference images from image-generation models, providing visual grounding of abstract language descriptions. In this paper, we provide empirical evidence suggesting our system can produce simple but tedious Blender editing sequences for tasks such as editing procedural materials and geometry from text and/or reference images, as well as adjusting lighting configurations for product renderings in complex scenes.
With the advent of large language models(LLMs) enhanced by the chain-of-thought(CoT) methodology, visual reasoning problem is usually decomposed into manageable sub-tasks and tackled sequentially with various external tools. However, such a paradigm faces the challenge of the potential "determining hallucinations" in decision-making due to insufficient visual information and the limitation of low-level perception tools that fail to provide abstract summaries necessary for comprehensive reasoning. We argue that converging visual context acquisition and logical reasoning is pivotal for tackling visual reasoning tasks. This paper delves into the realm of multimodal CoT to solve intricate visual reasoning tasks with multimodal large language models(MLLMs) and their cognitive capability. To this end, we propose an innovative multimodal CoT framework, termed Cantor, characterized by a perception-decision architecture. Cantor first acts as a decision generator and integrates visual inputs to analyze the image and problem, ensuring a closer alignment with the actual context. Furthermore, Cantor leverages the advanced cognitive functions of MLLMs to perform as multifaceted experts for deriving higher-level information, enhancing the CoT generation process. Our extensive experiments demonstrate the efficacy of the proposed framework, showing significant improvements in multimodal CoT performance across two complex visual reasoning datasets, without necessitating fine-tuning or ground-truth rationales. Project Page: https://ggg0919.github.io/cantor/ .