Abstract visual reasoning based on algebraic methods
Abstract Extracting high-order abstract patterns from complex high-dimensional data forms the foundation of human cognitive abilities. Abstract visual reasoning involves identifying abstract patterns embedded within composite images, considered a core competency of machine intelligence. Traditional...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-01-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-86804-3 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832571797105541120 |
|---|---|
| author | Mingyang Zheng Weibing Wan Zhijun Fang |
| author_facet | Mingyang Zheng Weibing Wan Zhijun Fang |
| author_sort | Mingyang Zheng |
| collection | DOAJ |
| description | Abstract Extracting high-order abstract patterns from complex high-dimensional data forms the foundation of human cognitive abilities. Abstract visual reasoning involves identifying abstract patterns embedded within composite images, considered a core competency of machine intelligence. Traditional neuro-symbolic methods often infer unknown objects through data fitting, without fully exploring the abstract patterns within composite images and the sequential sensitivity of visual sequences. This paper constructs a relation model with object-centric inductive biases, learning end-to-end multi-granular rule embeddings at different levels. Through a gating fusion module, the model incrementally integrates explicit representations of objects and abstract relationships. The model incorporates a relational bottleneck method from information theory, separating the input perceptual information from the embeddings of abstract representations, thereby restricting and differentiating feature processing to encourage relational comparisons and induce the extraction of abstract patterns. Furthermore, this paper bridges algebraic operations and machine reasoning through the relational bottleneck method, extracting common patterns of multi-visual objects by identifying invariant sequences within the relational bottleneck matrix. Experimental results on the I-RAVEN dataset demonstrate a total accuracy of 96.8%, surpassing state-of-the-art baseline methods and exceeding human performance at 84.4%. |
| format | Article |
| id | doaj-art-75a411253eec4df9ae3c3f9d45314816 |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-75a411253eec4df9ae3c3f9d453148162025-02-02T12:19:13ZengNature PortfolioScientific Reports2045-23222025-01-0115111010.1038/s41598-025-86804-3Abstract visual reasoning based on algebraic methodsMingyang Zheng0Weibing Wan1Zhijun Fang2School of Electronic and Electrical Engineering, Shanghai University of Engineering ScienceSchool of Electronic and Electrical Engineering, Shanghai University of Engineering ScienceSchool of Computer Science and Technology, Donghua UniversityAbstract Extracting high-order abstract patterns from complex high-dimensional data forms the foundation of human cognitive abilities. Abstract visual reasoning involves identifying abstract patterns embedded within composite images, considered a core competency of machine intelligence. Traditional neuro-symbolic methods often infer unknown objects through data fitting, without fully exploring the abstract patterns within composite images and the sequential sensitivity of visual sequences. This paper constructs a relation model with object-centric inductive biases, learning end-to-end multi-granular rule embeddings at different levels. Through a gating fusion module, the model incrementally integrates explicit representations of objects and abstract relationships. The model incorporates a relational bottleneck method from information theory, separating the input perceptual information from the embeddings of abstract representations, thereby restricting and differentiating feature processing to encourage relational comparisons and induce the extraction of abstract patterns. Furthermore, this paper bridges algebraic operations and machine reasoning through the relational bottleneck method, extracting common patterns of multi-visual objects by identifying invariant sequences within the relational bottleneck matrix. Experimental results on the I-RAVEN dataset demonstrate a total accuracy of 96.8%, surpassing state-of-the-art baseline methods and exceeding human performance at 84.4%.https://doi.org/10.1038/s41598-025-86804-3Abstract patternsInductive biasesEnd-to-endObject-centric |
| spellingShingle | Mingyang Zheng Weibing Wan Zhijun Fang Abstract visual reasoning based on algebraic methods Scientific Reports Abstract patterns Inductive biases End-to-end Object-centric |
| title | Abstract visual reasoning based on algebraic methods |
| title_full | Abstract visual reasoning based on algebraic methods |
| title_fullStr | Abstract visual reasoning based on algebraic methods |
| title_full_unstemmed | Abstract visual reasoning based on algebraic methods |
| title_short | Abstract visual reasoning based on algebraic methods |
| title_sort | abstract visual reasoning based on algebraic methods |
| topic | Abstract patterns Inductive biases End-to-end Object-centric |
| url | https://doi.org/10.1038/s41598-025-86804-3 |
| work_keys_str_mv | AT mingyangzheng abstractvisualreasoningbasedonalgebraicmethods AT weibingwan abstractvisualreasoningbasedonalgebraicmethods AT zhijunfang abstractvisualreasoningbasedonalgebraicmethods |