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..
See through the Dark: Learning Illumination-affined Representations for Nighttime Occupancy Prediction
Authors: Yuan Wu, Zhiqiang Yan, Yigong Zhang, Xiang Li, Jian Yang
NeurIPS 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on both real and synthetic datasets demonstrate the superior performance of LIAR under challenging nighttime scenarios. The source code and pretrained models are available here. |
| Researcher Affiliation | Academia | 1PCA Lab , Nanjing University of Science and Technology 2National University of Singapore 3Nankai University |
| Pseudocode | No | The paper describes methods and uses diagrams like Figure 2, but does not contain explicit pseudocode or algorithm blocks. |
| Open Source Code | Yes | The source code and pretrained models are available here. Source code and pretrained models are released for peer research. |
| Open Datasets | Yes | We evaluate our method on both real and synthetic nighttime scenarios. (1) Occ3Dnu Scenes [38] includes 700 scenes for training and 150 for validation, with annotations spanning a spatial range of -40m to 40m along both the X and Y axes, and -1m to 5.4m along the Z axis. (2) nu Scenes-C [51] is a synthetic benchmark that introduces eight types of data corruptions, each applied at three intensity levels to the validation set of nu Scenes. |
| Dataset Splits | Yes | Occ3Dnu Scenes [38] includes 700 scenes for training and 150 for validation... Notably, the training and validation sets include 84 and 15 real-world nighttime scenes, respectively. |
| Hardware Specification | Yes | Our implementation is based on MMDetection3D [4], and experiments are conducted on 4 NVIDIA Ge Force RTX 4090 GPUs. |
| Software Dependencies | No | The paper mentions "MMDetection3D [4]" and "Adam W optimizer [29]" but does not specify version numbers for these or other key software components. |
| Experiment Setup | Yes | During training, we employ the Adam W optimizer [29] with a learning rate of 2 10 4, training each model for 24 epochs. ... Finally, the total training loss Lt is formulated as: Lt = αLce + βLsem scal + γLgeo scal, (15) where α, β and γ are hyper-parameters and we empirically set to 10, 0.2, and 0.2, respectively. |