Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..
AF-UMC: An Alignment-Free Fusion Framework for Unaligned Multi-View Clustering
Authors: Bohang Sun, Yuena Lin, Tao Yang, Zhen Zhu, Zhen Yang, Gengyu Lyu
NeurIPS 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on various datasets demonstrate that our AF-UMC exhibits superior performance against other state-of-the-art methods. |
| Researcher Affiliation | Collaboration | Bohang Sun1,2, Yuena Lin1, Tao Yang3, Zhen Zhu4,5, Zhen Yang1, Gengyu Lyu1 1College of Computer Science, Beijing University of Technology 2Engineering Research Center of Intelligent Perception and Autonomous Control, Ministry of Education 3Idealism Beijing Technology Co., Ltd. 4School of Computer Science and Technology, Zhejiang Sci-tech University, China 5KEYI College, Zhejiang Sci-tech University, China |
| Pseudocode | Yes | Algorithm 1 The Training Process of AF-UMC. |
| Open Source Code | Yes | Additionally, the code is provided in the supplementary materials. |
| Open Datasets | Yes | Datasets. We employ ten widely-used multi-view datasets for comparative studies, which includes six small-scale datasets of Caltech7-5 [4], Handwritten [26], Scene [5], Caltech102-5 [4], Hdigit [1], Aloi [15] and four large-scale datasets NUSWIDEOBJ [16], Noisy MNIST [24], Cifar10 [42], Youtube Face [7]. |
| Dataset Splits | No | The paper uses ten widely-used multi-view datasets for comparative studies but does not provide specific details on how these datasets are split into training, validation, and test sets. It mentions using 'mini-batches' for training but no explicit split percentages or counts. |
| Hardware Specification | Yes | All experiments are conducted on the same machine with the Intel(R) Xeon(R) Gold 6148 2.40GHz CPU, 8 Ge Force RTX 3090 GPUs, and 512GB RAM. |
| Software Dependencies | No | AF-UMC trains 50 epochs on mini-batches of size 256 by using Adam optimizer [11] with a learning rate of 0.0003 in Py Torch [23] framework. |
| Experiment Setup | Yes | AF-UMC trains 50 epochs on mini-batches of size 256 by using Adam optimizer [11] with a learning rate of 0.0003 in Py Torch [23] framework. The hyperparameters γ and λ are set to 1 and 1, respectively. The encoder Ev and decoder Dv are respectively formulated by MLPs with dimensions {Dv, 500, 500, 2000, 512, c} and {d, 2000, 500, 500, Dv}, where the activation function is Re LU. |