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..
HyPlaneHead: Rethinking Tri-plane-like Representations in Full-Head Image Synthesis
Authors: Heyuan Li, Kenkun Liu, Lingteng Qiu, Qi Zuo, Keru Zheng, Zilong Dong, Xiaoguang Han
NeurIPS 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In this section, we conduct comprehensive qualitative and quantitative experiments on full-head image synthesis to demonstrate that our hy-plane representation is well-suited for rendering from any viewpoint. Our comparative analysis includes tri-plane, tri-grid, spherical tri-plane from EG3D, Pano Head, Sphere Head respectively, and various hy-plane variants and settings for ablation study, all trained on our dataset and pipeline. |
| Researcher Affiliation | Collaboration | Heyuan Li SSE, CUHK (Shenzhen) EMAIL Kenkun Liu SSE, CUHK (Shenzhen) EMAIL Lingteng Qiu Tongyi Lab, Alibaba Inc. EMAIL |
| Pseudocode | No | The paper describes methods and formulas but does not contain a clearly labeled 'Pseudocode' or 'Algorithm' block. |
| Open Source Code | Yes | We will open-source our code after the paper is accepted. In the supplementary material, we also submit key code related to the hy-plane representation and the near-equalarea warping strategy. |
| Open Datasets | Yes | We follow Pano Head and Sphere Head, using a training set that includes FFHQ Niemeyer and Geiger (2021), Celeb A Liu et al. (2018), LPFF Wu et al. (2023b), Wild Head Li et al. (2024), K-Hairstyle Kim et al. (2021), and a 6K in-house dataset of large-pose head images processed with Sphere Head s toolbox. |
| Dataset Splits | No | The paper states: "All training images are 512 512 in resolution and augmented with horizontal flips." and "The entire training process spans 25 million images." It also mentions using "50K real and synthetic samples" for FID calculation. However, it does not provide specific train/validation/test splits (e.g., percentages or exact counts) for the datasets used in the experiments. |
| Hardware Specification | Yes | All experiments are trained on eight NVIDIA V100 GPUs with a batch size of 32. |
| Software Dependencies | No | The paper does not provide specific version numbers for software dependencies such as libraries, frameworks, or programming languages (e.g., Python, PyTorch, CUDA versions). |
| Experiment Setup | Yes | All experiments are trained on eight NVIDIA V100 GPUs with a batch size of 32. We follow Pano Head and Sphere Head, using a training set that includes FFHQ Niemeyer and Geiger (2021), Celeb A Liu et al. (2018), LPFF Wu et al. (2023b), Wild Head Li et al. (2024), K-Hairstyle Kim et al. (2021), and a 6K in-house dataset of large-pose head images processed with Sphere Head s toolbox. All training images are 512 512 in resolution and augmented with horizontal flips. The entire training process spans 25 million images. |