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..
Temporal Smoothness-Aware Rate-Distortion Optimized 4D Gaussian Splatting
Authors: Hyeongmin Lee, Kyungjune Baek
NeurIPS 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments demonstrate the effectiveness of our method, achieving up to 91 compression compared to the original Ex4DGS model while maintaining high visual fidelity. These results highlight the applicability of our framework for real-time dynamic scene rendering in diverse scenarios, from resource-constrained edge devices to high-performance environments. |
| Researcher Affiliation | Collaboration | Hyeongmin Lee Twelve Labs EMAIL Kyungjune Baek Sejong University EMAIL |
| Pseudocode | No | The paper describes methods using mathematical formulations and textual descriptions but does not include any explicitly labeled 'Pseudocode' or 'Algorithm' blocks with structured steps. |
| Open Source Code | Yes | The source code is available at https://github.com/Hyeongmin LEE/RD4DGS. |
| Open Datasets | Yes | To validate the effectiveness of the proposed method, we conduct various experiments on two real-world datasets, including Neural 3D Video (N3V) [34] and Technicolor [35]. |
| Dataset Splits | No | To validate the effectiveness of the proposed method, we conduct various experiments on two real-world datasets, including Neural 3D Video (N3V) [34] and Technicolor [35]. Because the proposed method is built upon Ex4DGS [18], we match the evaluation protocol, it is originally adopted by the previous work [36]. Specifically, we use the entire set of N3V and the five scenes of Technicolor (Birthday, Fabien, Painter, Theater, and Train). |
| Hardware Specification | Yes | All experiments are conducted on an NVIDIA RTX 3090 GPU. |
| Software Dependencies | No | The paper mentions using Ex4DGS as a baseline and various compression techniques but does not specify version numbers for any software libraries or frameworks (e.g., PyTorch, CUDA, Python). |
| Experiment Setup | Yes | To systematically analyze performance at varying compression strengths, we define six compression levels (Levels 1 6), adjusting pruning hyperparameters as follows: for Gaussian point pruning, λGSprune is set to [0.05, 0.02, 0.01, 0.005, 0.002, 0.0005]; for SH coefficient pruning, λSHprune is set to [0.5, 0.2, 0.1, 0.05, 0.02, 0.005], respectively. |