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..
Wukong's 72 Transformations: High-fidelity Textured 3D Morphing via Flow Models
Authors: Minghao Yin, Yukang Cao, Kai Han
NeurIPS 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive quantitative and qualitative evaluations demonstrate that WUKONG significantly outperforms state-of-the-art methods, achieving superior results across diverse geometry and texture variations. |
| Researcher Affiliation | Academia | Minghao Yin1 Yukang Cao2 Kai Han1 1Visual AI Lab, The University of Hong Kong 2S-Lab, Nanyang Technological University EMAIL EMAIL EMAIL |
| Pseudocode | No | The paper describes the methods in detailed text and uses flowcharts (e.g., Figure 2) but does not include any explicit section or block labeled 'Pseudocode' or 'Algorithm'. |
| Open Source Code | Yes | Project page: https://visual-ai.github.io/wukong |
| Open Datasets | Yes | Following the evaluation protocol from (Sun et al., 2024), we use the SHREC07 (Temerinac et al., 2007) dataset and report performance using Dirichlet energy (Ezuz et al., 2019) and Coverage (Huang and Ovsjanikov, 2017) metrics. To evaluate the method s generalization to real 3D data, we conducted experiments using the Headspace dataset (Dai et al., 2020), which contains high-quality 3D face scans along with corresponding rendered RGB images. |
| Dataset Splits | Yes | Following the evaluation protocol from (Sun et al., 2024), we use the SHREC07 (Temerinac et al., 2007) dataset |
| Hardware Specification | Yes | We conduct experiments on an NVIDIA A100 GPU. |
| Software Dependencies | No | DINOv2 (Oquab et al., 2023) and CLIP (Radford et al., 2021) are used for image and text feature extraction. Both the structure and texture flow-based generator Φgeo, Φtex contain 21 cross-attention blocks... We use the free-support Wasserstein barycenter solver from the POT library (ot.lp.free_support_barycenter (Lindheim, 2023)). Justification: The paper mentions several software components and libraries, but it does not specify concrete version numbers for any of them. For example, it lists 'DINOv2', 'CLIP', and 'POT library', but not 'DINOv2 version X.Y' or 'POT library version Z.W'. |
| Experiment Setup | Yes | Both the structure flow and texture flow models are implemented using rectified flow with 25 sampling steps each. The Classifier-Free Guidance (CFG) strength is set to 3. For shape interpolation, we compute the token-wise barycenter using the free-support Wasserstein barycenter implemented by ot.lp.free_support_barycenter (Lindheim, 2023). We set the maximum number of optimization iterations to 100, and the convergence threshold (stop criterion) to 1 10 5. We conduct experiments on an NVIDIA A100 GPU. |