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..
Dynamic Focused Masking for Autoregressive Embodied Occupancy Prediction
Authors: Yuan Sun, Julio Contreras, Jorge Ortiz
NeurIPS 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We conduct experiments on three indoor occupancy prediction benchmarks: Occ-Scan Net [38], Embodied Occ-Scan Net [34], and their respective smaller variants, Occ-Scan Net-mini [34] and Embodied Occ-Scan Net-mini [34]. Following standard protocols [38, 34, 33], we report Scene Completion Intersection-over-Union (Io U) and mean Intersection-over-Union (m Io U) across semantic classes. Local metrics are computed within the camera frustum of each frame, while global metrics evaluate performance over the union of explored regions, capturing the model s ability to maintain spatial consistency over time. While our primary focus is on indoor occupancy prediction, we further assess the generalization capability of our approach on outdoor scenarios using publicly available datasets. These additional experiments underscore the broader applicability of our method. Experimental details and additional evaluation results are provided in Appendix. |
| Researcher Affiliation | Academia | 1Rutgers, The State University of New Jersey New Brunswick, NJ, USA 08901 EMAIL |
| Pseudocode | No | The paper describes methods and processes in narrative text and with mathematical formulations but does not present any explicitly labeled pseudocode or algorithm blocks. For example, Section 3 describes the Multi-Scale Gaussian Encoder and Hierarchical Multi-Scale Refinement. |
| Open Source Code | No | Question: Does the paper provide open access to the data and code, with sufficient instructions to faithfully reproduce the main experimental results, as described in supplemental material? Answer: [No] Justification: The code is not yet published, but the supplementary material provides all details needed to reproduce our model and results. |
| Open Datasets | Yes | We conduct experiments on three indoor occupancy prediction benchmarks: Occ-Scan Net [38], Embodied Occ-Scan Net [34], and their respective smaller variants, Occ-Scan Net-mini [34] and Embodied Occ-Scan Net-mini [34]. |
| Dataset Splits | Yes | Question: Does the paper specify all the training and test details (e.g., data splits, hyperparameters, how they were chosen, type of optimizer, etc.) necessary to understand the results? Answer: [Yes] Justification: Yes, we provide details of the dataset splits and the hyperparameter settings in Appendix. |
| Hardware Specification | Yes | Question: For each experiment, does the paper provide sufficient information on the computer resources (type of compute workers, memory, time of execution) needed to reproduce the experiments? Answer: [Yes] Justification: Yes, we report these details in Appendix, including the type of compute hardware used. We also provide memory consumption and runtime statistics related information in Tables 6, 7, and Appendix. |
| Software Dependencies | No | The paper does not explicitly list specific software dependencies with version numbers in the provided text. While the NeurIPS checklist indicates details for reproduction are in the Appendix, the visible text does not specify software versions like Python, PyTorch, or CUDA versions. |
| Experiment Setup | Yes | Question: Does the paper specify all the training and test details (e.g., data splits, hyperparameters, how they were chosen, type of optimizer, etc.) necessary to understand the results? Answer: [Yes] Justification: Yes, we provide details of the dataset splits and the hyperparameter settings in Appendix. |